Development of a WebGIS-based Decision Support System ...

Development of a WebGIS-based Decision Support System for Facilitating the Adoption of Agricultural Best Management Practices

by

Kun Chen

A Thesis presented to

The University of Guelph

In partial fulfillment of requirements for the degree of

Doctor of Philosophy in

Geography, Environment and Geomatics

Guelph, Ontario, Canada © Kun Chen, May, 2019

ABSTRACT

DEVELOPMENT OF A WEBGIS-BASED DECISION SUPPORT SYSTEM FOR

FACILITATING THE ADOPTION OF AGRICULTURAL BEST MANAGEMENT PRACTICES

Kun Chen Advisor(s):

University of Guelph, 2019 Wanhong Yang

Agricultural best management practice (BMP) adoption has the benefits of controlling

and reducing agricultural non-point source pollution. To facilitate the adoption, adequate

information needs to be provided to farmers and conservation managers to improve their

understanding on BMPs and support their decision making on BMP adoption. By utilizing

information and communication technologies, this study introduces the design and

implementation of a WebGIS-based decision support system to fulfill the information needs of

farmers and conservation managers for BMP adoption.

In the first step, this study develops an information model that conceptualizes information

communications within the BMP adoption process. The information model specifies the

information content for communications as well as defines how information could be generated

and communicated to meet the information needs of farmers and conservation managers, which

are classified into public information and BMP planning information based on the accessibility

of information.

Based on the information model, this study designs a WebGIS-based decision support

system for facilitating agricultural BMP adoption which includes three subsystems: the public

subsystem for supporting communications of landscape conditions and BMP educational

information, the BMP planning subsystem for supporting communications of BMP planning

information, and the administration subsystem for supporting administrative tasks including

monitoring the use of the BMP planning subsystem by farmers and conservation managers.

Based on the system design, a prototype of the WebGIS-based decision support system is

developed for the Gully Creek watershed, which is a representative watershed in southwestern

Ontario with active agricultural BMP implementation activities. The system prototype is then

evaluated by two methods: evaluation by direct use and evaluation during demonstration. The

evaluation by direct use identifies violence to usability principles, while the evaluation during

demonstration focuses on evaluating user task and information of the system. The results from

the two evaluation methods are coded into the evaluation measures and aggregated for

conducting an assessment of the system usability. The results show that the evaluators are overall

satisfied with the system design and functionalities. Several suggestions on further improvements

to the system are also provided.

iv

ACKNOWLEGEMENTS

I would like to acknowledge the Knowledge Translation and Transfer (KTT) Funding

Program of Ontario Agri-food Innovation Alliance for providing funding to my PhD research. I

would also like to acknowledge a list of people who were instrumental in supporting my PhD

research.

First of all, I would like to express my gratitude to Dr. Wanhong Yang for his continuous

support and encouragement in my Ph.D. journey. Without his support, I cannot accomplish it. To

me, he is not merely an advisor, but also a mentor for my life that steers me through all the

difficulties and challenges in my past four years of study. His skillful guidance, innovative ideas

and patience are greatly appreciated.

I would like to thank my committee members – Dr. John Lindsay and Dr. Songnian Li

who contributed to various discussions that helped to shape this research.

I would also like to thank Mari Veliz of Ausable Bayfield Conservation Authority, Darryl

Finnigan, Kevin McKague, Ross Kelly, Dr. Oswald Zachariah, Elin Gwen, Elin Gwyn, Tieghan

Hunt of Ontario Ministry of Agriculture, Food and Rural Affairs, Dr. Pradeep Goel of Ontario

Ministry of the Environment, Conservation and Parks, Jo-Anne Rzadki of Conservation Ontario,

Rebecca Moore and Shannon Brown of Ontario Agri-food Innovation Alliance, and Dr.

Bronwynne Wilton of Wilton Consulting Group for their valuable inputs and support.

Moreover, I would like to thank my colleagues Hui Shao, Yongbo Liu, Michael Tennant,

Peter Neill, Lane Buryta, Scott Schau, Gihan Sooriyabandara, and Mostafa Ghiyasvand for

v

supporting the WebGIS-based decision support system development and testing. It is a great

pleasure working with them and I appreciate their ideas and help.

At last, but not the least, I would like to acknowledge the patience and understanding

from my family.

vi

TABLE OF CONTENTS

ABSTRACT .................................................................................................................................... ii ACKNOWLEGEMENTS ............................................................................................................... iv TABLE OF CONTENTS ............................................................................................................... vi LIST OF TABLES ...................................................................................................................... viii LIST OF FIGURES ...................................................................................................................... ix

Chapter 1 Introduction ............................................................................................................ 1 1.1 Problem statement ............................................................................................................. 1 1.2 Purpose and objectives ...................................................................................................... 5 1.3 Thesis overview ................................................................................................................ 6

Chapter 2 Literature Review ................................................................................................... 7 2.1 Watershed management and planning for agricultural BMPs .......................................... 7 2.2 Farm economic and watershed hydrologic modelling for supporting agricultural BMP adoption ................................................................................................................................. 16 2.3 WebGIS for supporting agricultural BMP adoption ....................................................... 22 2.4 Research gaps .................................................................................................................. 32

Chapter 3 An Information Model for Agricultural BMP Adoption .................................. 34 3.1 The process of agricultural BMP adoption ..................................................................... 34 3.2 Developing an information model for the agricultural BMP adoption process .............. 38 3.3 Summary ......................................................................................................................... 47

Chapter 4 System Architecture and Design of the WebGIS-based Decision Support System for Facilitating Agricultural BMP Adoption .......................................................... 49

4.1 Task-oriented design of the WebGIS-based decision support system for facilitating agricultural BMP adoption .................................................................................................... 49 4.2 Integrating WebGIS and watershed modelling tools ...................................................... 53 4.3 The subsystems of the WebGIS-based decision support system for facilitating agricultural BMP adoption .................................................................................................... 56 4.4 The modules of the three subsystems ............................................................................. 59 4.5 The components of the system modules ......................................................................... 63 4.6 Summary ......................................................................................................................... 80

Chapter 5 A Prototype of the WebGIS-based Decision Support System for the Gully Creek Watershed .................................................................................................................... 82

5.1 Study area ........................................................................................................................ 82 5.2 The prototype development for the Gully Creek watershed ........................................... 84 5.3 Summary ....................................................................................................................... 117

Chapter 6 Evaluating the WebGIS-Based Decision Support System .............................. 118 6.1 Usability evaluation ...................................................................................................... 118 6.2 Evaluating the WebGIS-based decision support system using a qualitative approach 121 6.3 The evaluation results ................................................................................................... 131 6.4 Summary ....................................................................................................................... 135

vii

Chapter 7 Conclusion ........................................................................................................... 137 7.1 Summary ....................................................................................................................... 137 7.2 Research contributions .................................................................................................. 140 7.3 Future study .................................................................................................................. 143

REFERENCES .......................................................................................................................... 146 APPENDIX A ............................................................................................................................. 172 APPENDIX B ............................................................................................................................. 176 APPENDIX C ............................................................................................................................. 178 APPENDIX D ............................................................................................................................ 182 APPENDIX E ............................................................................................................................. 186

viii

LIST OF TABLES

Table 3-1: Information needs for BMP adoption .......................................................................... 39 Table 3-2: Information communication technologies and tools ................................................... 42 Table 4-1: Tasks of system modules in the public subsystem ...................................................... 59 Table 4-2: Tasks of system modules in the BMP planning subsystem ......................................... 61 Table 4-3: Tasks of system modules in the administration subsystem ......................................... 62 Table 5-1: Software for system development ............................................................................... 88 Table 5-2: System localization checklist ...................................................................................... 89 Table 5-3: Various combinations of water quantity/quality effects and BMP costs .................. 105 Table 5-4: The parameters for producing the BMP policy/management information ................ 108 Table 6-1: Measures for evaluating the WebGIS-based decision support system ...................... 124 Table 6-2: Key user tasks for evaluation by direct use ............................................................... 128 Table 6-3: System design for evaluation during demonstration ................................................. 131

ix

LIST OF FIGURES

Figure 2-1: Understanding the roles of GIS-based decision support system in planning ............. 22 Figure 3-1: Key stakeholders and their roles in the information communication process for BMP adoption ......................................................................................................................................... 35 Figure 3-2: Key questions driving the BMP adoption .................................................................. 39 Figure 3-3: The information sub-model for field characteristics, environmental concerns, and BMP adoption ............................................................................................................................... 44 Figure 3-4: The information sub-model for agri-environmental policy and BMP related technical knowledge ..................................................................................................................................... 45 Figure 3-5: The information sub-model for BMP planning .......................................................... 46 Figure 4-1: Task-oriented design of the WebGIS-based decision support system for facilitating agricultural BMP adoption ............................................................................................................ 52 Figure 4-2: Loose coupling of WebGIS and the integrated economic-hydrologic model ............ 54 Figure 4-3: Loose coupling of the WebGIS and the optimization model ..................................... 55 Figure 4-4: Subsystems in the WebGIS-based decision support system for facilitating agricultural BMP adoption ............................................................................................................ 58 Figure 4-5: Component diagram of the “Information sharing site” module ................................. 64 Figure 4-6: Component diagram of the “Public information center” module .............................. 66 Figure 4-7: Component diagram of the “Access control” module ............................................... 67 Figure 4-8: Component diagram for BMP scenario creation ........................................................ 68 Figure 4-9: Component diagram for BMP scenario development ................................................ 69 Figure 4-10: Component diagram for BMP scenario evaluation .................................................. 70 Figure 4-11: Component diagram for BMP evaluation result exploration ................................... 71 Figure 4-12: Component diagram for BMP scenario comparison ................................................ 72 Figure 4-13: Component diagram for scenario optimization ........................................................ 73 Figure 4-14: Component diagram for optimization result exploration ......................................... 74 Figure 4-15: Component diagram of the “Discussion” module .................................................... 76 Figure 4-16: Component diagram of the “Report” module .......................................................... 77 Figure 4-17: Component diagram of the “User registration” module .......................................... 78 Figure 4-18: Component diagram of the “System monitoring” module ....................................... 79 Figure 5-1: The Gully Creek Watershed in Southern Ontario, Canada ........................................ 83 Figure 5-2: The welcome webpage of the public subsystem ........................................................ 91 Figure 5-3: The interface of the information sharing site ............................................................. 92 Figure 5-4: The form for uploading information on field characteristics and BMP adoption ...... 93 Figure 5-5: The interface of the public information center ........................................................... 94 Figure 5-6: The login webpage of the BMP planning subsystem ................................................. 95 Figure 5-7: The form for creating a “What if” BMP scenario ...................................................... 96 Figure 5-8: The WebGIS interface for developing a "What if" BMP scenario ............................ 97 Figure 5-9: BMP assignments in the Gully Creek watershed ....................................................... 98 Figure 5-10: The WebGIS interface for scenario evaluation result presentation and exploration 99 Figure 5-11: The WebGIS interface for scenario comparison .................................................... 101 Figure 5-12: Differences in net return between the “What if” and the baseline scenario .......... 102 Figure 5-13: Differences in total phosphorus between the "What if" and the baseline scenario 103 Figure 5-14: Cost-effectiveness of BMPs on total phosphorus reduction .................................. 104

x

Figure 5-15: The WebGIS interface for the “Policy/Management” module .............................. 106 Figure 5-16: The default range of BMP policy/management constraints ................................... 107 Figure 5-17: The interface for exploring environmental policy/management information ........ 109 Figure 5-18: The interface for exploring economic policy/management information ............... 110 Figure 5-19: The interface for supporting communications between farmers and conservation managers ..................................................................................................................................... 111 Figure 5-20: Scenario reports in HTML(Left) and PDF(Right) formats .................................... 112 Figure 5-21: The webpage after login to the administration subsystem ..................................... 113 Figure 5-22: Tables for displaying the usage information of the BMP planning subsystem ..... 114 Figure 5-23: Communication network for displaying communication information among farmers and conservation managers ......................................................................................................... 115 Figure 5-24: Interaction within the communication network ..................................................... 116 Figure 5-25: The user registration form ...................................................................................... 117 Figure 6-1: The updated Information System Success Model .................................................... 123 Figure 6-2: Evaluation by direct use and evaluation during demonstration ............................... 127

1

Chapter 1 Introduction

1.1 Problem statement

As the world population increases, global agriculture must address the challenge of

supplying the escalating demand for agricultural production. While the intensification of

agriculture by use of high-yielding crop varieties, fertilization, irrigation, and pesticides has

contributed to significant increases in food production over the recent decades (Mastson et al.,

1997), it has meanwhile contributed to non-point source water pollution, which is increasingly

disruptive to the freshwater system (McCoy, Chao, & Gang, 2015). For several decades, tenable

evidence has indicated that excessive input of Nitrogen (N) and Phosphorous (P) can change

water chemistry with subsequent eutrophication and food web modification (Udeigwe et al.,

2011), N and P leached from agricultural fields can contaminate the groundwater (Böhlke, 2002),

and toxic chemicals entering the food supply can cause unexpected diseases to animals and

human beings (Lu et al., 2015).

The establishment of agri-environmental programs reflects a current trend of agricultural

development towards a more sustainable approach (Garnett et al., 2013). By providing financial

incentives to farmers to implement best management practices (BMPs) such as conservation

tillage, nutrient management, cover crop, and water and sediment control basin (WASCoB),

these programs aim to meet agricultural production goals while preventing the excessive

sediment and nutrient loadings into water bodies. The adoption of BMPs is the core of these agri-

environmental programs and many scientific efforts have been made to examine the rationale for

BMP adoption. For example, several studies have conducted meta-analyses to examine the

2

relationships between farm and farmer characteristics and farmer’s BMP adoption (Prokopy et

al., 2008; Baumgart-Getz, Prokopy, & Floress, 2012), while other studies have focused on

analyzing the perception of innovations and farmer’s attitude change towards adoption (Reimer,

Weinkauf, & Prokopy, 2012; Trujillo-Barrera, Pennings, & Hofenk, 2016).

The extensive studies in agricultural BMP adoption contribute to our understanding on

agricultural BMP adoption, and one important implication from those studies is the significance

of information in driving the BMP adoption process and supporting BMP adoption decisions. As

Baumgart-Getz et al. (2012) suggested, the success of agricultural BMP adoption relies on the

provision of adequate and appropriate information to stakeholders. Feather and Amacher (1994)

also noted that information can be a primary reason for widespread adoption of BMPs. Indeed, as

key stakeholders for BMP adoption, farmers and conservation managers require information to

develop understanding on environmental and economic effects of BMPs (Prager et al., 2012).

Farmers also need information to improve their environmental awareness and technical

knowledge about BMPs (Rogers, 1995).

Given the significance of information for BMP adoption, agri-environmental programs

have widely adopted a participatory approach and various efforts have been made to promote

information communications among stakeholders. For example, the Canada-Ontario

Environmental Farm Plan organized workshops and face-to-face consultations to help

stakeholders with their BMP adoption decisions (Smithers & Furman, 2003). The European

Union’s rural development policy also required State Members to develop stakeholder

involvement and partnership programs for developing and implementing agri-environmental

policies (Prager & Freese, 2009). However, significant barriers still exist. As reported by

3

Nxumalo and Oladele (2013), some key obstacles to participation in agri-environmental

programs include lack of technical knowledge, communication, sense of stewardship, and funds.

Several key challenges must be addressed to facilitate stakeholder participation in agri-

environmental programs and improve their communications for BMP adoption. Firstly, as

adopters of BMPs, farmers need to be motivated to participate in the process of communication.

This requires them to realize their stewardship role and understand BMP characteristics.

Secondly, to improve the effectiveness of communications among stakeholders, a common

ground among stakeholders, particularly farmers and conservation managers, should be built

based on mutual understanding on the economic costs, environmental benefits and cost-

effectiveness of BMPs (Prager et al., 2012). Finally, new methods of communication need to be

developed. In addition to traditional communications such as face-to-face consultations which

require physical presence, the new communication methods should provide more flexibility and

convenience through utilizing information technologies.

Information on economic costs, environmental benefits and cost-effectiveness of BMPs is

essential for supporting effective communications among the stakeholders for BMP adoption.

Farmers can utilize the information to understand BMP characteristics and develop plans for

BMP adoption on their farms. Conservation managers can use the information to spatially target

land parcels for BMP implementation in order to minimize the economic costs and/or maximize

the environmental benefits. In recent years, efforts have been made to develop integrated

economic-hydrologic modelling systems to understand the trade-offs between costs and benefits

of BMPs (Srivastava et al., 2002; Yang et al., 2003; Turpin et al., 2005). For example, Qi and

Altinakar (2011) proposed a conceptual framework to support multi-objective decision making

4

for BMP allocation. The integrated approach coupled the hydrologic model AnnAGNPS, the

channel network model CCHE1D and an economic model to estimate both the economic costs

and environmental benefits of land use plan alternatives.

However, while very useful, these integrated modelling systems are typically complex. It

is a challenge to make the modelling information accessible and understandable by stakeholders.

Many studies have suggested that coupling geographic information system (GIS) with the

integrated modelling systems can improve the accessibility of information through spatial data

visualization techniques (Jayakrishnan et al., 2005; Olivera et al., 2006; Liu, Bralts, & Engel,

2015; Karki et al., 2017; Shao et al., 2017; Jang, Ahn, & Kim, 2017). However, those GIS

interfaces are largely confined to desktop applications and used mostly by experts.

Given both the managerial challenge to improve information communications for BMP

adoption and the technical challenge to extend GIS applications for improved information

accessibility, a WebGIS-based decision support system can be developed to overcome these

challenges. As an extension to the traditional desktop-based GIS, the WebGIS-based decision

support system empowers multi-users to address spatial decision-making tasks using GIS

functions and information and communication technologies (ICTs) (Sieber, 2006). With the

advent of the Internet, the WebGIS-based decision support system enables access to spatial

information and services without time and place constraints (Kingston et al., 2000).

Designing a WebGIS-based decision support system for facilitating BMP adoption

requires the system to satisfy the various information needs of stakeholders. The information

needs of stakeholders, particularly farmers and conservation managers, can be identified based

on their roles in the process of BMP adoption. However, how to utilize information technologies

5

and communication tools to generate, communicate and present the information to stakeholders

still remains in question. As such, an information model needs to be developed to address how

different information and communication technologies can be utilized to facilitate information

communications for BMP adoption. The information model should identify information content

for communications and also define the details of information communications in terms of

information sources, targets and channels.

1.2 Purpose and objectives

The purpose of the research is to develop a WebGIS-based decision support system for

facilitating the adoption of agricultural BMPs. Specifically, the research has four interrelated

objectives:

1) Develop an information model to conceptualize the information communication

process for agricultural BMP adoption.

2) Design a WebGIS-based decision support system to facilitate information

communications for agricultural BMP adoption. The WebGIS-based decision support system

should provide stakeholders with easy access to relevant information for BMP adoption. The

WebGIS-based decision support system should also improve communications among

stakeholders to support their consensus-building and decision-making for BMP adoption.

3) Develop a prototype of the WebGIS-based decision support system for facilitating the

adoption of agricultural BMPs.

6

4) Evaluate the system prototype to understand whether and how the system supports

user tasks and meet the information needs by stakeholders

1.3 Thesis overview

In the thesis, Chapter 2 reviews relevant literature and identifies research gaps. Chapter 3

analyzes the agricultural BMP adoption process and develops an information model for

conceptualizing information content and communication process for BMP adoption. In Chapter

4, the design of the WebGIS-based decision support system for facilitating agricultural BMP

adoption is presented. In Chapter 5, a prototype of the WebGIS-based decision support system

for a study area is developed. The system interface, user interactions, and information

presentations are illustrated. In Chapter 6, the usability evaluation of the system is presented. In

the Chapter 7, the conclusions and future research are discussed.

7

Chapter 2 Literature Review

The literature review covers three research subjects. Firstly, watershed management and

planning for agricultural BMPs is reviewed. The review focuses on understanding the agri-

environmental programs and corresponding information needs of stakeholders for BMP

adoption. Secondly, watershed modelling of agricultural BMPs is reviewed. The review

examines the applications of watershed hydrologic and farm economic models and integrated

hydrologic-economic models for providing valuable information on BMP costs, effectiveness

and cost-effectiveness. Lastly, applications of WebGIS for supporting collaborative tasks are

reviewed. The review helps to illustrate the role of the decision support system in planning and

identify the strengths and limitations of WebGIS design and implementation for supporting

agricultural BMP adoption.

2.1 Watershed management and planning for agricultural BMPs

This section reviews agri-environmental programs and discusses the related challenges.

Information needs of farmers and conservation managers for BMP adoption are also reviewed.

2.1.1 Agricultural BMPs

Agricultural best management practices (BMPs) are measures to mitigate agricultural

non-point water pollution (Logan, 1993). Two types of agricultural BMPs are structural and non-

structural BMPs. The structural BMPs involve stationary and permanent facilities to prevent or

reduce the discharge of pollutants (Ackerman & Stein, 2008). Examples of structural BMPs

include water and sediment control basins, vegetated filter strip and riparian buffers. The non-

8

structural or land management BMPs work by changing farming behaviour through government

regulations, persuasion and/or economic instruments (US EPA, 1999). Some non-structural

BMPs include cover crops, fertilizer management, and conservation tillage.

Non-structural or land management BMPs have two approaches to mitigate the non-point

water pollution. The first approach is to reduce inputs of harmful substances on agricultural

fields; the second approach is to control erosion and runoff. Nutrient management is a

representative practice that fits into the first approach. It reduces the pollution by reducing the

use of fertilizers and manure on agricultural fields (Havlin et al., 1999). Conservation tillage and

cover crop are examples of land management practices that use the second approach. They

mitigate the pollution by preventing erosion and reducing the transport of sediment and nutrients

on the fields (Dabney, 1998; Holland, 2004).

Agricultural BMPs improve water quality, but they incur economic costs. The economic

costs involve different cost categories such as construction costs, production costs and

opportunity costs. The construction costs are expenditures for building BMP structures or

facilities such as water and sediment control basins (Weiss et al., 2007). The production costs

involve, for example, the costs for purchasing fertilizers, the expenditures for using and

maintaining agricultural machinery, and the labour effort of learning and implementing the

farming technologies (Veith et al., 2003). The opportunity costs are the profit losses from

changes in farm operations (Veith et al., 2003).

Planning for BMP implementation requires an integrated evaluation from both

environmental and economic aspects. Given that the environmental benefits (i.e. pollution

reduction) and economic costs of BMPs vary for different BMP combinations and locations,

9

understanding the cost-effectiveness of BMPs is essential for BMP adoption. As indicated by

Shao et al. (2017), the cost-effectiveness information is important for conservation managers to

develop BMP implementation plan and farmers to make BMP adoption decisions.

2.1.2 Agri-environmental programs

The adverse effects of intensive agricultural activities have led to the establishment of

various agri-environmental programs. By providing financial incentives, agri-environmental

programs aim to engage farmers to adopt BMPs to meet environmental objectives. For example,

the Conservation Reserve Program (CRP) under the U.S. Department of Agricultural (USDA)

offered different types of payments, including both one-time and annual payments for land

retirement by farmers. The average federal cost for CRP could reach up to $2 billion per year

(Stubbs, 2014). As a part of the Environmental Quality Incentive Program (EQIP), the Great

Lake Restoration Initiative (GLRI) also provided approximately 300 million per year, up to

$2.56 billion from FY2010 to FY2017, to assist farmers to use scientifically proven conservation

practices to protect watershed and shorelines from non-point source pollution (“GLRI funding”,

n.d.). In Canada, the Environmental Farm Plan (EFP) offered financial assistance for farmers to

implement BMPs to reduce potential damage to water bodies from agricultural activities

(Morrison & FitzGibbon, 2014). The Rural Water Quality Program (RWQP) in the Grand River

Conservation Authority also provided grants in the Ontario regions of Waterloo, Oxford and

Wellington to compensate farmers up to 50 to 100 percent costs of the selected BMPs (Dupont,

2010).

With the significant amount of conservation investment, it is important for agri-

environmental programs to achieve cost-effectiveness. For this purpose, a strategy is to target the

10

investment to various combinations of BMPs and land parcels that yield the greatest

environmental benefit per dollar of cost (Engel et al., 2008). Successful targeting requires

information on the cost-effectiveness of agricultural BMPs to improve the investment efficiency.

In recent years, complex watershed modelling has been increasingly applied to evaluate the BMP

cost-effectiveness and generate landscape process-based results. Shao et al. (2017), for example,

developed an integrated hydrologic-economic modelling system for BMP evaluation and policy

design.

Information on the cost-effectiveness of agricultural BMPs helps conservation managers

to design the implementation of agricultural BMPs to maximize the environmental benefits.

Because farmers are agricultural BMP adopters, the information also needs to be communicated

to farmers in order to develop a mutual understanding between conservation managers and

farmers towards BMP adoption. To facilitate this communication process, agri-environmental

programs adopted a participatory approach. The Canada-Ontario Environmental Farm Plan, for

example, offered a peer-to-peer support to assist farmers to develop BMP action plans (Smithers

& Furman, 2003). Stakeholder workshops were also held to support farmers to understand the

BMP adoption process (Prager & Freese, 2009).

Despite significant efforts to facilitate BMP adoption, farmer participation can be still

hampered by various constraints. Smithers and Furman (2003) suggested that the participation is

strongly associated with several farmer, farm and program characteristics. In particular, they

noted that the education level of famers and their awareness of existing environmental issues can

impose a significant influence on their participation. This conclusion is also supported by Luzar

and Diagne (1999) who suggested that farmers’ environmental attitude is important for active

11

participation. In addition, Breetz et al. (2005) indicated that farmers’ trust on the program is an

important factor for determining their willingness to participate. Furthermore, from a

management perspective, Falconer (2000) suggested that participation can be hindered by the

transaction cost such as visiting and reporting requirements.

To facilitate farmers’ participations in agri-environmental programs, information and

communication technologies (ICTs) offer great potential. Chapman et al. (2002) suggested that

ICTs can improve the access to information and create linkages among farmers and conservation

managers for information sharing. Glendenning and Ficarelli (2012) also suggested that by

maintaining a two-way information communication among conservation managers and farmers,

ICTs could provide farmers with quick access to agricultural BMP information and also allow

conservation managers to reach out to farmers with timely and accessible support. In addition,

Richardson (2006) showed that ICTs can improve farmers’ knowledge and effectiveness of their

communications, thereby increase their chance of participation. Furthermore, Silva (2008)

revealed that ICTs, once designed and used appropriately, can vastly reduce the cost of

information communications and increase the likelihood of participation in agri-environmental

programs.

Incorporating ICTs into agri-environmental programming has become a current trend for

addressing agricultural BMP planning tasks. Improving information communications and

increasing farmers’ exposure to a variety of information such as the cost-effectiveness of BMPs,

BMP know-how knowledge and availability of financial incentives can support farmers’

engagement in agri-environmental programs and promote BMP adoption (Leach, Pelkey, &

Sabatier, 2002; Llewellyn, 2007). However, to make the best use of ICTs for agricultural BMP

12

adoption, several challenges still need to be addressed regarding social structure of

communications, information generations, and information quality.

2.1.3 Agricultural BMP adoption and related information needs

2.1.3.1 Information needs for farmers to adopt BMPs

Agricultural BMP adoption can be viewed as a process of information communications

wherein stakeholders collectively address environmental and economic concerns and finalize a

plan for implementing agricultural BMPs. To encourage BMP adoption, it is necessary to

provide relevant and adequate information to farmers to improve their knowledge and further

support their decision-making on BMP adoption. Thus, it is important to identify their

information needs for BMP adoption. In the past decades, extensive studies have been carried out

to examine the rationale for BMP adoption. Based on these studies, a variety of information

needs of farmers for adoption can be identified.

Farmers require information to understand why implementing BMPs is important and

necessary. Studies suggested that information on existing environmental problems can improve

farmers’ environmental awareness and hence adoption. From a diffusion perspective, Rogers

(1995) explained that being aware of existing environmental problems is important for famers to

realize a need for adoption. Based on the Theory of Planned Behavior (Ajzen, 1991), Kaiser et

al. (1999) also noted that an improved awareness of environmental problems is necessary for

farmers to develop an attitude towards those problems and an intention to adopt BMPs. By

conducting a survey in the tropical savannas of southern Australia to investigate farmers’

motivations to BMP adoption, Greiner and Gregg (2011) also found that being informed of

environmental issues presents an important reason for farmers’ adoption.

13

Farmers also require information to improve their knowledge on BMPs (McCown, 2002;

Reimer et al., 2012). Rogers (1995) stated that farmers should be educated with three types of

BMP knowledge including BMP concepts, functions, and implementation details. Specifically,

he explained that 1) knowledge on BMPs concepts and functions could help prevent farmers’

rejection and discontinuance of these practices and 2) understanding how to properly implement

BMPs could increase the probability of successful adoption of these practices. Many empirical

studies have shown that improving farmer’s understanding on BMP characteristics can be crucial

to support farmers’ adoption. Alonge and Martin (1995), for example, revealed that an improved

understanding on BMP characteristics such as complexity and compatibility can be an influential

factor for farmers’ decision making on adoption. Dietz et al. (2004) also suggested that educating

farmers on how BMPs contribute to mitigate non-point source pollution can be essential for

fostering the voluntary adoption of BMPs.

Moreover, farmers require information to support their decisions on BMP adoption.

Specifically, information should assist them to understand the costs and benefits of BMP

adoption. The BMP costs include installation cost, maintenance cost and the opportunity cost due

to yield loss, and the BMP benefits include financial benefits such as government compensation

and non-financial benefits such as environmental benefits (Atari et al., 2009). BMP costs and

benefits can vary due to various combinations of BMPs and the geographical locations where

they are applied.

Farmers are sensitive to BMP cost information and many studies have indicated that

BMP costs can have a substantial impact on adoption. In a survey to evaluate the efficiency of

the Environmental Farm Plan (EFP), Plummer et al. (2008) reported that nearly half of the

14

farmers identified implementation costs to be a major barrier to their undertaking of agricultural

BMPs. In examining the motivations of farmers’ participation in the Nova Scotia EFP program,

Atari et al. (2009) also found that majority of the farmers (53%) identified high BMP costs as a

dominant factor for preventing them from adoption. Moreover, by conducting a survey to

investigate farmers’ BMP adoption motivations, Greiner and Gregg (2011) reported the cost of

time and labour and the loss of productivity and/or profitability to be the main causes for non-

adoption.

While the BMP adoption introduces private costs, it also provides environmental

benefits. Studies revealed that information on environmental benefits could have a positive

influence on farmers’ adoption of BMPs. For example, Bultena and Hoiberg (1983) noted that

being aware of BMP’s environmental benefits is positively related to farmers’ conservation

behaviors. In a study to investigate farmers’ motivation for BMP adoption, Reimer et al. (2012)

also indicated that farmers identified soil conservation such as improved soil structure/fertility is

their major motivation for adoption. Moreover, Greiner and Gregg (2011) identified farmers’

awareness of improvements on environmental conditions to be a major motivation for BMP

adoption.

To evaluate alternative BMP plans and make adoption decisions, farmers require

integrated assessment of BMP costs and effectiveness. The cost-effectiveness of BMPs reflects

environmental benefits obtained per dollar of cost. Shao et al. (2017) suggested that information

on the cost-effectiveness of BMPs can provide a more efficient means for farmers to understand

the effects of BMPs on their fields. They also noted that the cost-effectiveness information is a

15

key factor for making a mutual agreement on BMP adoption between farmers and conservation

managers.

Several studies illustrated that communications among farmers can improve the

possibility of BMP adoption. Based on the diffusion theory, Rogers (1995) suggested that

information communications among farmers are important to induce farmers to adopt

agricultural practices because it improves the observability of the practices. He explained that

compared to mass media such as newspaper and magazine, opinions from other farmers can be

more effective and convincing to engage farmers to adopt. Similarly, Reimer et al. (2012)

indicated that farmers are more likely to adopt BMPs if they have observed positive economic

and environmental outcomes of adoption by others. Furthermore, Lubell and Fulton (2007), by

discussing the impact of local diffusion network on BMP adoption, noted that information

communications among farmers can improve social trust and norms and may induce cultural

change on BMP adoption.

2.1.3.2 Information needs of conservation managers for BMP implementation

Economic costs, environmental benefits and cost-effectiveness of BMPs at watershed and

regional scales are necessary for conservation managers to design BMP policy/management to

improve the efficiency of investment on BMP implementation. In the United States, several agri-

environmental programs, such as Conservation Reserve Program and Environmental Quality

Incentives Program, have noted that BMP cost-effectiveness information is essential for benefit-

cost targeting of BMPs (Claassen et al., 2008). In Canada, the OMAFRA also initiated programs

to understand cost-effectiveness of BMPs to facilitate targeting agricultural BMPs on the land

with greatest environmental benefits (Smithers & Furman, 2003). Shao et al. (2017) also noted

16

that the BMP cost-effectiveness is necessary for conservation mangers to design two types of

policies: 1) how to maximum the environmental benefits given a fixed investment and 2) how to

minimize the cost given a fixed target of environmental improvement.

2.2 Farm economic and watershed hydrologic modelling for supporting agricultural BMP adoption

In recent decades, farm economic and watershed hydrologic modelling has been

increasingly applied to support decision-making on agri-environmental programs and BMP

adoption. In particular, an integrated economic-hydrologic modelling approach has been widely

adopted to generate information on the cost-effectiveness of BMPs. This section reviews the

applications of farm economic and watershed hydrologic models as well as the integrated

hydrologic-economic systems for BMP evaluation.

2.2.1 Farm economic modelling

Various studies have developed farm economic modelling to understand the impact of

conservation practices on farming economics. Yiridoe et al. (2000), for example, used farm

economic modelling to estimate economic costs and cropping net return of tillage systems. The

production costs included both farm input cost and machinery cost. The farm input cost was

modelled as a function of several input variables such as seed and fertilizer, and the machinery

cost was estimated based on equipment usage such as oil and lubrication, repair and

maintenance. The farm gross return was calculated by multiplying the yields by the market price.

The net return was calculated by subtracting the production cost from the farm gross return. By

simulating and comparing different tillage systems at two sites in Ontario of Canada, they found

that due to higher machinery-related cost, no-till system has less production cost than

17

conventional tillage and other reduced tillage systems. They also indicated that while the reduced

tillage systems obtained similar results on the average net farm returns, the choice of preferred

tillage system should depend on the climate and soil type.

In a recent study, Yang et al. (2016) developed integrated economic-hydrologic

modelling to examine wetland restoration scenarios in the South Tobacco Creek watershed in

Manitoba, Canada. In the farm economic modelling, the costs of wetland restoration included

opportunity cost and administration and engineering cost. The opportunity cost was estimated

based on a yield function and an average historical crop price. This study compared economic

costs under four levels of wetland restoration scenarios with different subset of wetlands for

restoration. They found that to achieve the same benefits of total phosphorus (TP) reduction, full

wetland restoration was associated with the minimum average economic cost. As shown in their

modelling results, the average economic cost of full restoration was $132.4 ha/yr at a TP

reduction level of 1.9 kg/ha/yr. The second partial restoration scenario costed more at $135.9

ha/yr while resulting in a less TP reduction level at 1.7 kg/ha/yr.

Shao et al. (2017) conducted an integrated economic-hydrologic assessment of

agricultural BMPs in a study watershed in Ontario. In the farm economic model, production

variables included machinery, seed and fertilizer. The net return was calculated by subtracting

production costs from revenue (crop yields multiplied by prices). The costs of land management

BMPs including conservation tillage, cover crop and nutrient management were estimated as net

return differences under the conventional baseline scenario with no BMPs and a BMP scenario.

The cost of Water and Sediment Control Basins (WASCoBs) was estimated based on

engineering costs associated with earthwork and outlet installation.

18

2.2.2 Watershed hydrologic modelling

Watershed hydrologic modelling provides information to understand the environmental

benefits of BMPs. In recent decades, many watershed models have been applied for evaluating

environmental effects of agricultural BMPs. According to the literature (Xie et al., 2015), several

frequently used models include the Soil and Water Assessment Tool (SWAT) (Arnold et al.,

1998), Agricultural Nonpoint Source (AGNPS) (Young et al., 1989), Annualized Agricultural

Nonpoint Source (AnnAGNPS) (Bingner & Theurer, 2003) and Hydrological Simulation

Program-FORTRAN (HSPF) (Bicknell et al., 2005).

Characterizing BMPs within these watershed models is typically fulfilled by changing

model input or adjusting parameters or implementing specific modules to reflect changes of the

hydrologic process due to BMPs (Xie et al., 2015). Arabi et al. (2008) introduced

characterization of several BMPs within the SWAT model. For example, contour farming was

simulated through modifications of SCS curve number and USLE practice factor, and cover crop

was simulated by scheduling a crop rotation within a year in the model input to simulate cover

crop management. In Yang et al. (2013), a Water and Sediment Control Basin (WASCoBs)

module was developed and incorporated into the SWAT as the SWAT has no modules

specifically designed to simulate the water quality and quantity effects of WASCoBs.

The SWAT (Arnold et al., 1998) has been widely used by a large number of studies to

evaluate BMP effects in mitigating NPS pollution. Parajuli et al. (2008), for example, quantified

the effects of vegetative filter strips in the 950-km2 Wakarusa watershed in northeast Kansas. As

a result, a 63% decrease in sediment was reported. Lee et al. (2010) also evaluated the reduction

in NPS pollution by applying vegetative filter strips, riparian buffer system and fertilizing control

19

for a 1.21 km2 agricultural watershed. The modelling results indicated a 16–25% reduction in

sediment, 5–37% reduction in total nitrogen (TN), and 6–41% reduction in total phosphorus (TP)

respectively. Furthermore, Maharjan et al. (2016) evaluated the BMPs of split fertilizer

application (SF), cover crop cultivation (CC), and the combination (SFCC) at a watershed scale.

They found that the SF scenario reduced nitrate pollution and sediment compared to single

fertilizer application, the application of the CC scenario reduces both sediment and nitrate

loadings, and the SFCC showed the highest positive effect on reducing sediment and nitrate.

2.2.3 Integrated economic-hydrologic modelling

Given that both environmental benefits and economic costs of BMPs are important for

making BMP adoption decisions, a number of studies have been carried out to develop integrated

economic-hydrologic modelling systems to evaluate cost-effectiveness of multiple BMPs. Yang

and Weersink, (2004), for example, developed an integrated economic-hydrologic modelling

approach to examine cost-effective targeting of riparian buffers in the Canagagigue Creek

watershed in southern Ontario. The economic returns of crop production were calculated by

estimating the revenue, production costs and quasi-rent of the production. The sediment

abatement benefits were obtained by combining the AnnAGNPS watershed model and a field-

scale Vegetation Filter Strip (VFS) model. By simulating the economic costs at different levels

of sediment abatement, they reported crop return losses for achieving the sediment abatement

goals. Specifically, the average costs for achieving 10%, 30% and 50% sediment abatement

goals were $175, $227 and $306/ha, respectively. Moreover, they suggested that compared to

implementing the riparian with a fixed width, allowing for buffer strips of different sizes can

increase the cost-effectiveness significantly.

20

Yang et al. (2005) also developed the integrated economic-hydrologic modelling

approach for spatial targeting of conservation tillage to improve water quality and carbon

retention benefits. The cost-effectiveness of conservation tillage was obtained from integrated

results from three models. The Crop Budget model was used to estimate the revenue, production

cost, and net return of the practice, the SWAT model was used to simulate the hydrologic

process and calculate the sediment abatement benefits, and the Century solid organic model was

used to estimate the carbon retention benefits from implementing conservation tillage. A GIS

interface was developed to facilitate the modelling process by preparing modelling input and

visualizing the integrated modelling results. Based on the cost-effectiveness of conservation

tillage, the modelling system was capable of preforming BMP optimization and suggesting BMP

design policies. Based on evaluating several sediment abatement and carbon retention goals, they

concluded that using the targeting strategy for implementing conservation tillage can obtain joint

benefits of improving water quality and retaining soil organic carbon while minimizing

economic costs.

Moreover, Kelly et al. (2018) developed an integrated economic-hydrologic approach for

targeting water retention pond BMP to reduce phosphorous runoff in the Lake Winnipeg

watershed of southern Manitoba. Two components of the economic costs were estimated

including construction cost and opportunity cost. The reduction in total phosphorus was

estimated by the SWAT model. After obtaining the environmental benefits and economic costs

of each water retention pond, the study compared three targeting strategies (i.e. cost targeting,

benefit-maximum targeting and benefit-cost targeting) and revealed that under the same

phosphorus reduction objective, the benefit-cost targeting strategy prioritized the locations for

21

water retention ponds with the highest cost-effectiveness in terms of providing the greatest level

of environmental benefits.

There are also other studies on integrated economic-hydrologic modelling for BMP cost-

effective analysis. Osei et al. (2000) used a Comprehensive Economic and Environmental

Optimization Tool (CEEOT) to estimate phosphorous reduction and associated cost under

different manure management scenarios. Ghebremichael et al. (2013) developed an integrated

modelling framework by combining the SWAT model and a farm-level economic model to

evaluate changes in the cost of various BMPs. By applying the integrated Universal Soil Loss

Equation (USLE) and economic model to a 1,014 ha watershed in Rickingham County of

Virginia, Veith et al. (2004) examined optimal BMP allocation for improving BMP cost-

effectiveness with respect to sediment reduction. The modelling results showed that BMP

implementation can result in up to $49 per ka/ha sediment reduction. Gitau et al. (2004)

integrated the SWAT model and an economic model to simulate dissolved P and cost of different

BMP scenarios at a 300-ha farm within the Town Brook watershed of Delaware County in

Pennsylvania. With a target of 60% dissolved phosphorous reduction, the model identified the

most cost-effective scenario with suggested site-specific BMP combinations.

Integrated economic-hydrologic modelling has been increasingly studied in recent

decades to evaluate the impact of BMPs on environmental benefits and economic costs.

However, a long-standing challenge has been how to make such modelling systems and the

associated information accessible to stakeholders to support their decision-making. To address

the challenge, the integrated modelling system outputs needs to be communicated to meet user

22

information needs for BMP adoption. Furthermore, the system needs to be user-friendly with an

interface that is easy to learn and manipulate.

2.3 WebGIS for supporting agricultural BMP adoption

2.3.1 GIS-based decision support system (DSS)

The GIS-based decision support system (DSS) has been increasingly developed in recent

decades. Malczewski (2004) suggested to understand the role of GIS-based decision support

system in planning from two types of rationality, named instrumental rationality and

communicative rationality (Figure 2-1). The instrumental rationality emphasizes spatial

reasoning and analysis as the core of planning while the communicative rationality suggests the

importance of communication support in planning for user engagement, conflict resolution, and

ultimately, consensus making.

Figure 2-1: Understanding the roles of GIS-based decision support system in planning (Based on Malczewski 2004)

From the instrumental perspective, GIS-based spatial analysis method can be classified

into two main categories: computer-assisted overlay mapping methods and multi-criteria

decision-making (MCDM) methods. The computer-assisted overlay mapping methods, such as

GIS-based Decision Support System

Instrumental rationality Communicative rationality

23

Boolean operations and weighted linear combination (WLC), address the suitability problems

through overlay analysis. As examples, Mbilinyi et al. (2007) developed an overlay mapping

method for identifying potential sites for rainfall harvesting. The overlaying analysis was

conducted by combining six map layers, including rainfall, soil depth, slope, soil texture,

landcover and drainage layers. Similarly, Qi and Altinakar (2011) developed an overlay mapping

analysis for estimating the flood damage in the area of Milledgeville, Georgia of the United

States. The flood damage was estimated based on one layer on land feature types and the other

layer on flood depth across the region. The flood depth-damage relationships on different land

feature types were developed using the data from field surveys and expert panel opinions.

The MCDM methods advance the computer-assisted overlay mapping methods by

implementing multi-criteria decision rules with mathematical programming algorithms. The

result of the MCDM methods can be largely associated with the accuracy and precision of the

input-data to the GIS-based multi-criteria procedure (Zhou and Civco, 1996). In recent decades,

the MCDM methods have been widely used to support various spatial decision-making tasks,

such as site selection, urban planning, and environmental management. As examples, Rinner and

Malczewski (2002) developed a decision support tool based on the Order Weighted Averaging

(OWA) for resort site selection. The OWA evaluated the suitability of resort sites based on two

sets of weights (i.e. criterion important weights and order weights) and was able to support

decisions based on the attitude of decision-makers (e.g. risks and trade-off). Sugumaran et al.

(2004) developed a GIS-based decision support system for environmental planning and

watershed management. In the system, a decision-making process was implemented to calculate

the environmental sensitivity index of watersheds using 13 weighted parameters, such as portion

of watershed area with slope greater than 15% and relative abundance of endangered species.

24

Van Haaren et al. (2011) also developed a GIS-based decision support system for siting wind

farms in the New York State. They suggested that the locations with the highest wind resources

are not always feasible sites for wind farm, and introduced a site selection tool to calculate the

site priority based on multiple economic, planning, environmental and ecological parameters,

such as electronic line and land costs, noise and visual impact, slope, bird habitat, and distance to

lakes and rivers.

From the communicative perspective, the GIS-based decision support systems facilitate

planning in two ways. Firstly, the decision support system facilitates consensus building through

modelling. As examples, Boroushaki and Malczewski (2010a) developed a participatory GIS and

used the fuzzy majority approach to obtain the group solution based on individual preferences. A

consensus measuring tool was also implemented based on the calculation of consensus measure

and proximity measure. The consensus measure showed the agreement among individual

preferences to the group solution and the proximity measure calculated how close the individual

preferences are to the group solution (Boroushaki & Malczewski, 2010b). Mekonnen and

Gorsevski (2015) developed a WebGIS-based decisions support system for siting wind farms

within Lake Eire area, Ohio. A voting system was integrated to allow participants to score wind

farm alternatives based on several parameters such as population density, distance from shore

and bird habitat. Participants’ scores then were aggregated using the Borda count to promote a

consensual solution.

Secondly, the decision support system can support consensus building by improving

information accessibility and communications among individuals. With advancement in internet-

based communication technologies, several studies have been conducted to understand the use of

25

WebGIS to improve the planning process (Kingston et al., 2000; Rinner, 2001; Elwood & Ghose,

2011). A detailed review of the WebGIS-based communication technologies is given in the

Section 2.3.2.

2.3.2 Information communications with WebGIS

The WebGIS emerges out of the GIS and enables access to spatial information and

services through the Internet without time and place constraints (Kingston et al., 2000).

Consequently, the WebGIS offers greater flexibility for two or more people to address collective

decision-making tasks using GIS functions (Elwood & Ghose, 2011). Given that agricultural

BMP adoption requires frequent communications among stakeholders, particularly farmers and

conservation managers, WebGIS can be an ideal tool to support adoption of agricultural BMPs

(Zhang et al., 2011).

WebGIS has been used to support a variety of information communication tasks

(Karatzas et al., 2000; Kingston, 2007; Sidlar & Rinner, 2007; Simão et al., 2009). In its simplest

form, the WebGIS offers a one-way communication and is mainly used as a platform to deliver

information to the public. In those systems, users are in a passive mode and can only extract

geographic information from the map. Some typical examples of this form of WebGIS include a

series of interactive environmental maps published by the Government of Canada for the public

to observe the distribution of environmental variables, such as air quality, water quality and

quantity (“Open maps,” n.d.). By delivering information to the public, such WebGIS plays an

important role in educating people and improving their knowledge. However, those systems

typically lack functionalities to support public communications and complex decision-making

tasks.

26

Some studies used WebGIS as an important means for collecting information. These

studies implemented the concept of volunteered geographic information (VGI) (Coleman et al.,

2009). VGI represents a shift in how geographic information is created and shared (Elwood et

al., 2012). In the VGI applications, participants of VGI applications are information producers

(Coleman et al., 2009). Kingston et al. (2000), for example, developed a WebGIS for supporting

local environmental decision-making. The system adopted an open-ended approach and allowed

members of the community to make comments on a public map. The public map enabled the

function to collect and share local knowledge for the future environmental development. Based

on a preliminary survey, the authors found that the public response to the system was positive.

Particularly, the respondents found the ability to type in comments useful. Similarly, Kingston

(2007) introduced a WebGIS for supporting local policy decision-making. The system provided

an online mapping interface for citizens to report environmental problems. The interface also

supported policy officers to monitor problems in real time and target resources to problem

locations. As a conclusion, Kingston (2007) stated that such online interactive mapping tools

provided an effective means to improve communication efficiency and citizen participation in

the decision-making process.

Volunteered geographic information offers a low-cost mechanism for the acquisition of

geographic information; however, it prompted concerns with regard to its quality. Flanagin and

Metzger (2008) defined the quality of VGI as the extent to which the participants provide their

personal inputs honestly and accurately. To assure the quality of VGI, Goodchild and Li (2012)

introduced two main approaches, i.e. the crowd-sourcing and social approaches. The crowd-

sourcing approach relies on the ability of the crowd to validate and correct the errors that an

individual might make, while the social approach relies on a group of trusted individuals as gate-

27

keepers to validate the information. Moreover, Goodchild and Li (2012) emphasized that the

validation of VGI should rely on the broad body of geographic knowledge. As an example, they

showed that “Joe’s Café” can be mislocated to a historic area which factually contains no

business.

More complex WebGIS offers two/multiple-way communications and allows the public

to initiate and engage online discussions and disseminate information. Rinner (2001) developed

the argumentation map model to represent spatially related discussion. The model allowed geo-

referencing argumentation elements (e.g. comments) to geographic objects and vice versa. Based

on the argumentation map model, Keßler (2004) implemented a software called Argumap. Sidlar

and Rinner (2007) later evaluated an Argumap prototype for supporting participants to discuss

planning ideas and concerns in the University of Toronto. Based on participants’ feedback on the

Argumap prototype, Sidlar and Rinner (2007) found that the Argumap was useful for supporting

spatially related discussions. They also received several recommendations for the prototype

improvement. Those recommendations included map tool tips, multimedia content support, and

help menu.

After the introduction of the argumentation map model, many GIS-based online

discussion applications were developed. Tang (2006), for example, developed a GIS-based

online discussion forum (GeoDF) for campus planning. The system supported online discussions

and allowed participants to send proposed plans to other participants within a specific distance.

As an improvement to the argumentation map, the GeoDF focused on the spatial context to

support discussions. Tang (2006) defined the concept of spatial context to include spatial and

text components. The spatial component includes map extent, visible map layers, sketches, and

28

annotations. Moreover, Brent et al. (2010) developed a WebGIS called “MapChat tool” to

facilitate information communications among participants. The MapChat tool allowed

participants to draw new features on the map and add comments to the selected map features.

Because all comments were spatially referenced, the MapChat tool allowed participants to

examine comments based on locations, and spatial relevance of comments. Brent et al. (2010)

concluded that the MapChat tool supported a self-directed and sustainable participatory approach

for information collection and can benefit the process of local knowledge acquisition.

Some studies used the argumentation map in conjunction with Spatial Decision Support

System (SDSS) to support collaborative decision-making tasks. For example, Simão et al. (2009)

developed a WebGIS for wind energy planning. The system combined a Multi-Criteria Spatial

Decision Support System (MC-SDSS) and an argumentation map. The MC-SDSS evaluated

wind farms and classified them into three pre-defined categories (i.e. Recommended,

Acceptable, and Non-acceptable). The argumentation map rendered the classification result and

allowed users to make spatially referenced comments associated with the wind farm sites. To

manage the comments, the system used a discussion forum with a tree structure to maintain the

logic chain of arguments. Based on system evaluation, the authors noted that the system

improved interactions among farmers and provided a learning opportunity to facilitate users to

understand the complexity of wind-farm siting.

With the development of ICTs, many communication tools have been incorporated into

WebGIS to facilitate synchronous or asynchronous communications. As examples, Karatzas et

al. (2000) presented project APNEE (Air Pollution Network for Early warning and online

information Exchange in Europe). In the project, they used a WebGIS to promote dissemination

29

of air quality information. The system allowed citizens to access air quality information from

different information channels. Specifically, the system provided Short Message Service (SMS)

and voice mail for citizens to receive air quality information on their mobile devices. The system

also provided email function for citizens to receive newsletters and active notifications. By

offering different communication channels, the WebGIS facilitated better communication to

improve mutual understanding between citizens and city authorities. Bugs et al. (2010) also

discussed the use of email as an effective tool for asynchronous information transfer and opinion

sharing. Moreover, Butt and Li (2012) developed a web-based, collaborative GIS prototype to

support public participation. The prototype examined the use of ICTs and showed a great

potential of ICTs to improve public participation. Specifically, the prototype provided a shared

map for participants to explore geographic context of tasks and incorporated a GIS-based

discussion forum to manage geo-referenced comments posted by participants. The prototype also

prepared a virtual public meeting interface to facilitate map-based communications among

participants. The virtual public meeting interface provided an array of facilitation tools for

participants to interact with the shared map (e.g. select features, add annotations, and draw

graphics).

The various communication methods and tools have provided great potential for WebGIS

to support information communication tasks. However, current technological competency

doesn’t necessarily lead to the success of stakeholder participation and problem solving. Brown

(2012) and Brown and Kyttä (2014) reviewed the use of WebGIS for regional and environmental

planning and identified several concerns that need to be addressed for applying WebGIS to

specific information communication tasks. Firstly, it is important to identify system users and

their information needs. The users may include decision makers, implementers, affected

30

individuals, interested observers, or the general public (Schlossberg & Shuford, 2005).Secondly,

there is a need to increase the use of WebGIS. They suggested that incentives such as

sponsorship, education and better system design are some possible approaches to engage the

public for WebGIS participation. Finally, efforts need to be made to identify and control threats

to spatial data quality, particularly for WebGIS portals where users are granted with open access

to contribute and disseminate information.

2.3.3 GIS-based decision support system for facilitating agricultural BMP adoption

With the advances in GIS technologies and availability of digital spatial data, progress

has been made to develop GIS-based decision support systems for watershed management and

planning. In such decision support systems, watershed modelling tools were often employed as

MCDM components for supporting spatial analysis and reasoning. The functions of GIS in those

decision support systems have been reviewed by many researchers. As indicated by Goodchild,

Parks and Steyaert (1993), GIS has been employed to improve the accessibility of the modelling

system through facilitating the organization and visualization of spatial data and providing

spatial analysis functions to assist modelling tasks. Jayakrishnan et al. (2005) also noted that GIS

helped assemble the required spatial data from GIS coverages, create necessary model input files

efficiently, and enable water resources professionals to study large watershed systems with

significant savings in time and cost. Moreover, Olivera et al. (2006) indicated that interfacing

GIS with watershed modelling system has greatly improved the efficiency and easiness of using

the modelling systems. As one of the leading GIS companies, Environmental System Research

Institute (ESRI) published an ArcGIS-SWAT tool for watershed modelling (Olivera et al., 2006).

The GIS in the ArcGIS-SWAT tool automates multiple watershed modelling tasks, including

31

watershed delineation, classification of Hydrologic Response Unit, preparation of modelling

input files and visualization of modelling results.

Many GIS-based decision support systems have been designed and developed in recent

years for analyzing the cost-effectiveness of BMPs and supporting BMP planning. Rao (2007),

for example, integrated the SWAT into the ArcIMS to evaluate the environmental benefits of the

Conservation Reserve Program in the Beaver River watershed in Oklahoma of the United States.

Shao et al. (2017) also developed a GIS-based decision support system for BMP evaluation and

policy making. These GIS-based decision support systems were able to generate essential

information for support BMP adoption, however, they were mostly desktop-based and mainly

used by experts (Liu, Bralts, & Engel, 2015; Karki et al., 2017; Jang, Ahn, & Kim, 2017). Those

systems have the potential to be further developed to improve the accessibility of modelling

functions and facilitate information communications on modelling results among stakeholders.

The rationale behind this is that the decision support system should fulfill both instrumental and

communicative roles in the process of agricultural BMP adoption (Malczewski, 2004).

WebGIS provides a promising platform to improve information communications among

stakeholders for agricultural BMP adoption. Some WebGIS systems have been applied to

address environmental management issues. Reviewing those systems can be helpful for

designing a WebGIS-based decision support system for facilitating agricultural BMP adoption.

Luchette and Crawford (2008) developed a WebGIS to visualize and publish pollution

information for citizen-based monitoring. In the system, an online interactive map was provided

to the public for exploring water quality data by hydrologic unit. Werts et al. (2012) also

developed an online web mapping interface that enabled users to both submit pollution

32

information and explore pollution pattern. By incorporating social media, the interface allowed

pollution information sharing among communities. Moreover, Ahmed et al. (2017) built a mobile

application called “My City, My Environment”. With mobility support, the system enabled users

to access information at any time and from anywhere. A web service was built in the application

for users to report pollution incidents. As the system was open to any user, user trust and privacy

protection were a major concern (i.e. whether the data was trustworthy, and users might feel

unsafe to report incidents). It would be necessary to adopt a credential system that requires users

to register in the server before using the reporting web service.

Some of the functionalities from those WebGIS systems have the potential to be used for

supporting agricultural BMP adoption. However, new functions and modules need to be

designed based on the information communication process for BMP adoption.

2.4 Research gaps

Information communication is important for stakeholders to reach a consensus on

agricultural BMP adoption. Many studies on agricultural BMP adoption have provided insights

on information needs of stakeholders, particularly farmers and conservation managers. However,

it is necessary to improve the understanding on information communication process among

stakeholders, which include the role of stakeholders in the communication process, their

information needs, and their communication tasks. With the potential of information

technologies to fulfill the information needs of stakeholders, it is particularly necessary to

improve the knowledge about the use of various information technologies to support information

communications among stakeholders for agricultural BMP adoption.

33

Integrated economic-hydrologic modelling systems have been increasingly developed to

generate information on economic cost, environmental benefits and cost-effectiveness of

agricultural BMPs. Engaging farmers for BMP adoption requires such information to be readily

available and communicated to them. Conservation managers also need the information to

conduct spatial targeting of agri-environmental programs. However, most of the integrated

modelling systems are desktop-based and primarily used by experts. These systems have the

potential to be extended to improve user access to information in BMP adoption process.

GIS-based decision support systems have been developed for agricultural BMP adoption.

Those systems were able to provide functions for evaluating the effects of agricultural BMPs, but

they also need to fulfill the communicative role for the BMP adoption. The WebGIS has become

a promising tool to support information communication tasks. Various WebGIS-based

applications have been developed to address environmental issues and some of the functionalities

of those systems can be utilized to further develop a WebGIS-based decision support system for

facilitating agricultural BMP adoption. The WebGIS-based decision support system needs to be

developed based on understanding the information communication process for agricultural BMP

adoption.

34

Chapter 3 An Information Model for Agricultural BMP Adoption

This chapter develops an information model that will be used to design a WebGIS-based

decision support system for facilitating agricultural BMP adoption. The information model

conceptualizes information communications within the BMP adoption process and derives

information needs of farmers and conservation managers. More specifically, it defines

information contents and also how information should be generated and communicated.

Scientific researchers, farmers and conservation managers are key stakeholders within the

process. Scientific researchers are environmental modellers that generate information on BMP

cost-effectiveness, farmers are BMP adopters, and conservation mangers are BMP

policy/management designers and adoption facilitators. Within the information model,

information technologies such as GIS and ICTs are examined to address various information

needs of the stakeholders.

3.1 The process of agricultural BMP adoption

3.1.1 Stakeholders and their roles in the information communication process for BMP adoption

Figure 3-1 illustrates key stakeholders and their roles in the information communication

process for BMP adoption. The key stakeholders include scientific researchers, conservation

managers, and farmers. Scientific researchers identify agri-environmental problems. Scientific

researchers also design agricultural BMPs and develop the integrated economic-hydrologic

models for BMP evaluation (Yang & Weersink, 2004; Yang, Sheng, & Voroney; 2005; Yang et

al., 2016). The integrated economic-hydrologic models play an important role to generate the

35

information on private economic costs, public environmental benefits and cost-effectiveness of

BMPs.

Figure 3-1: Key stakeholders and their roles in the information communication process for BMP adoption

Conservation managers have the responsibility to design BMP policy/management (e.g.

set up environmental targets) and allocate investments to incentivize and facilitate BMP

adoption. To support conservation managers to fulfill their roles, scientific researchers provide

conservation managers with information to educate farmers on BMP concepts and

implementation. Scientific researchers also provide conservation managers with information on

BMP costs, benefits and cost-effectiveness to support them to design BMP policy/management.

The cost-effectiveness information helps conservation managers to determine the minimum

investment to achieve specific environmental targets and BMP locations. The information also

helps conservation managers to maximize environmental benefits under financial constraints

(Shao et al., 2017).

Scientific researchers

Conservation managers

Farmers

Integrated economic-hydrologic models for BMP

evaluation

Agricultural best management practices

(BMP)

Agri-environmental problems

Educate/Transfer information

Educate/Persuade

Develop

Design

Identify

EvaluateAddress

Adopt

36

Farmers are adopters of BMPs. According to literature (Prokopy et al, 2008; Knowler &

Bradshaw, 2007), BMP adoption is related to various factors such as farm characteristics (e.g.

slope, area and soil), farmer characteristics (e.g. education and age), and economic factors (e.g.

farmers’ income and commodity prices). A number of hypotheses have been tested regarding the

effects of those factors on farmers’ adoption of BMPs. However, due to physical, social and

economic differences across study areas, their conclusions on explaining the adoption rationale

vary (Knowler & Bradshaw, 2007). For example, Carlson et al. (1981) found no relationship

between age and adoption of conservation practices, whereas D’Souza et al. (1993) found that

younger farmers are more likely to adopt new technologies than older farmers. While

some researchers suggested a dominant role of financial considerations on adoption, Vanclay

(1992) reported that farmers do not necessarily act in economically rational ways because non-

financial factors such as stewardship attitude can also affect BMP adoption.

To promote farmers’ adoption of BMPs, conservation managers need to educate farmers

on BMP concepts and implementation. Conservation managers also need to persuade farmers to

adopt BMPs. The private economic costs, public environmental benefits, and cost-effectiveness

of BMPs are the most direct information to support farmers to make decisions on BMP adoption.

This information is essential to improve farmers’ understanding on the effects of BMPs (Rogers,

1995). This information is also important to support communications between farmers and

conservation managers to achieve a consensus on BMP adoption (Shao et al., 2017).

Furthermore, communications among farmers on their adoption experiences can be important to

improve the possibility of BMP adoption by farmers (Reimer et al., 2012).

37

3.1.2 Improving the process of BMP adoption

Agri-environmental programs support the BMP adoption process by providing

stakeholders with communication opportunities such as in-person consultations and workshops

(Smithers & Furman, 2003; Prager & Freese, 2009). This approach to communications, however,

has limitations. Firstly, planning those communication opportunities such as in-person

consultations and workshops may incur considerable costs. The high costs may affect the

efficiency of agri-environmental programs (Coggan et al., 2010). The costs such as travel costs

and time can also become a barrier for stakeholders to taking part in the process of

communications (Falconer, 2000; Smithers & Furman, 2003; Claassen et al., 2008). Secondly, as

the communication opportunities are generally planned at fixed times and locations and require

physical presence of stakeholders, those communication opportunities lack the flexibility to

sufficiently support the communication process. Considering that BMP adoption might require

frequent communications among stakeholders to develop a mutual understanding of BMPs and

address their concerns about BMP adoption (Rogers, 1995), an enhancement to the

communication system is necessary to improve stakeholders’ access to information and increase

their communications. As indicated by Ma et al. (2012), ample communication opportunities are

essential for farmers to obtain information to reduce their perceived risks of BMP adoption.

Advances in information technologies in recent decades have revolutionized the way of

communications and can provide great opportunities to overcome the limitations of current

approach to information communications for BMP adoption. On one hand, information

technologies can improve information provision to stakeholders. Information technologies allow

scientific researchers and conservation managers to publish information and modelling tools on

38

the Internet. Information technologies also allow farmers and conservation managers easily

access the information and modelling systems. Secondly, information technologies enable online

communications which offer a more flexible ways of communications among the stakeholders

for BMP adoption. The online communications can facilitate mainly two types of

communications for BMP adoption (Lubell & Fulton, 2007; Reimer et al., 2012). One type of

communications is between conservation managers and farmers. Conservation managers can

provide farmers with BMP program information and other information supports such as BMP

technical knowledge. Farmers also can request information from conservation mangers. The

other type of communications is among farmers.

3.2 Developing an information model for the agricultural BMP adoption process

This section discusses how an information model is developed to characterize

information communication process for the agricultural BMP adoption. The development of the

information model takes three steps. Firstly, the information needs by farmers and conservation

managers are analyzed. Their information needs are classified into two categories, public

information and BMP planning information, according to information characteristics. Secondly,

based on the information classification, the use of ICTs for different information needs is

analyzed. Finally, the information model is developed to characterize the information

communication process supported by information technologies.

3.2.1 Public and BMP planning information

Information is the key for stakeholders to work together to finalize a plan for

implementing BMPs. During the process of BMP adoption, farmers and conservation managers

39

need information to address several key questions including “What is the environmental

problem?”, “How to address the problem?”, “What are the environmental and economic effects

of BMPs?”, and “What is the optimal solution for BMP implementation based on cost-

effectiveness?” (Figure 3-2).

Figure 3-2: Key questions driving the BMP adoption

The information needs of farmers and conservation managers can be classified into two

categories: public information and BMP planning information (Table 3-1).

Table 3-1: Information needs for BMP adoption

Public information Field characteristics, environmental concerns, and BMP adoption

BMP related technical knowledge

Agri-environmental policies

BMP planning

information

Economic costs, environmental benefits and cost-effectiveness of BMPs

BMP policy/management subject to economic constraints or environmental targets

Communications between farmers and conservation managers

What is the environmental problem?

What is the optimal solution for BMP implementation

based on cost-effectiveness?How to address the problem?

What are the environmental and economic effects of

BMPs?

Information

40

Public information: The public information allows public access. Information contents of

the public information could include 1) farmers’ field characteristics, environmental concerns,

and BMP adoption, 2) BMP related technical knowledge, and 3) agri-environmental policies.

These information contents play different roles in facilitating BMP adoption. For example,

information on field characteristics, environmental problems, and BMP adoption can improve

farmers’ environmental awareness and motivation for landscape stewardship. Educating farmers

with BMP related technical knowledge would equip them with “know-how” of BMPs and

compatibility with the existing farming system. Information on agri-environmental policies can

help farmers to understand environmental regulation on farming operations and also economic

incentives for BMP adoption (Greiner & Gregg, 2011).

BMP planning information: The BMP planning information is field and farm specific and

can be only accessed by farmers and conservation managers. It aims to support BMP planning

activities such as BMP evaluation and implementation. Information contents of the BMP

planning information can include 1) economic costs, environmental benefits and cost-

effectiveness of BMPs, 2) BMP policy/management information such as environmental targets

for a watershed and financial constraints on BMP investment, and 3) communications between

farmers and conservation managers. Information on BMP economic costs, environmental

benefits and cost-effectiveness is necessary for farmers and conservation managers to evaluate

field-specific BMP effects for various “What if” BMP scenarios. The information on BMP

policy/management can be useful to support spatial targeting of BMPs to meet environmental

targets with minimized costs and to maximize environmental benefits under financial constraints

at a watershed scale. The communication between farmers and conservation managers is

necessary for them to achieve a mutual consensus on BMP adoption.

41

The differentiation of information into public information and BMP planning information

has implications on developing the WebGIS-based decision support system. For the BMP

planning information, authorization mechanisms need to be built to control the information

access and communications when field and farm specific confidential information is involved.

Also, the differentiation of information can help select information communication tools and

technologies for different communication tasks.

3.2.2 Technologies and tools for information communications

Supporting information communications for BMP adoption needs to specify information

sources, information targets and information channels. Information sources are where

information is sourced or generated, information targets are where information is delivered or

received, and information channels define how information should be communicated between

information sources and targets. Some of the traditional information channels include newspaper,

television, radio and interpersonal networking (Westerman, 2008). However, with the

development of ICTs, channels for information communications have been revolutionized in

terms of their timeliness and efficiency (Westerman, 2008). Complementary to traditional

communication means, these information communication technologies and tools can be

integrated into the WebGIS-based decision support system to facilitate BMP adoption.

The information communication technologies and tools can be classified into two main

categories – tools for synchronous communications and tools for asynchronous communications

(Table 3-2). The tools for synchronous communications include chats, video conferencing,

interactive whiteboard and voice over IP. Chats can be described as online text conversations

that happen in real-time (Peris et al., 2002). The chats could involve two or more persons. The

42

two main ways of conducting chats include instant massager and web-based chats. Video

conferencing is mostly used for business collaborations. It enables immediate multi-point

meetings based on various time zones (Panteli & Dawson, 2001). Voice over IP is similar to

video conferencing; but instead of meeting in person, it is purely audio-based. At last,

whiteboards are popular online communication tools in education. They allow users to write,

draw and even collaborate with the help of an interface which simulates an actual physical

whiteboard (Kershner et al, 2010).

Table 3-2: Information communication technologies and tools

Synchronous Chats Video conferencing Voice over IP Interactive whiteboards

Asynchronous Public information portal Online forum Email

The synchronous communication tools are suitable for collaborative tasks requiring real-

time communications. However, as synchronous communications between farmers and

conservation managers can be challenging, the WebGIS-based decision support system for

facilitating agricultural BMP adoption can mainly utilize the tools for asynchronous

communications. Public information portal, as a collection of information from diverse

information sources, is suitable for delivering and disseminating the public information. Online

forums can be described as places where all users are allowed to post either comments or

questions (Janssen & Kies, 2005). Other users of the forums are permitted to reply to posts so as

to create online discussions. Online forums include discussion groups, discussion boards, and

bulletin boards. In online forums, discussion posts can be properly stored to be chronologically

43

or thematically sorted to form threads. In the context of BMP adoption, it could offer an effective

tool to organize the communications among farmers and conservation managers. Moreover,

electronic mail or email can be used to deliver important messages or files such as photographs

and files (Ebert & Shapiro, 2009).

3.2.3 Information modelling for agricultural BMP adoption

Information modelling for agricultural BMP adoption can be developed based on

information classification and examination of ICTs. Information modelling for public

information communications focuses on conceptualizing how public information should be

generated, compiled and disseminated to the public; while information modelling for BMP

planning information communications explains how BMP planning information is generated and

communicated between farmers and conservation managers for supporting BMP adoption (i.e.

BMP evaluation and policy/management design).

3.2.2.1 Information modelling for public communications

Information sharing among farmers about their fields and BMP adoption

Information sharing among farmers can have a positive impact on BMP adoption. Figure

3-3 shows the process of information sharing among farmers. Farmers share information on field

characteristics, environmental concerns, and BMP adoption by submitting annotations which are

linked to their fields. An annotation can record information such as crop type, soil type, soil

quality, erosion, and land management practices in its linked field. Based on the links between

annotations and fields, a WebGIS interface can be used to present the information. The WebGIS

interface can also provide a communication platform for farmers to explore annotations

submitted by other farmers.

44

Figure 3-3: The information sub-model for field characteristics, environmental concerns, and BMP adoption

Agri-environmental policy and BMP related technical knowledge

Agri-environmental policy and BMP related technical knowledge are public information

for facilitating BMP adoption. Figure 3-4 shows the information sub-model for communicating

information on agri-environmental policy and BMP related technical knowledge. Information on

agri-environmental policies can be collected from webpages. BMP related technical knowledge

can be obtained from webpages and BMP technical documentation. Conservation managers can

collect information from these different information sources and import the information to the

public information portal. The public information portal provides easy access to the information

and has the ability to consolidate and deliver information in an organized way (Murray, 2002;

Zhang et al., 2016). It also allows farmers to explore information in a systematic manner such as

finding information by keywords and facilitates farmers’ self-learning process.

Farmers Annotations

WebGIS

share

are displayed onis explored by

Crop typeSoil type

Land management practices

Soil quality

Erosion

Fields

own are linked to

45

Figure 3-4: The information sub-model for agri-environmental policy and BMP related technical knowledge

3.2.2.2 Information modelling for BMP planning

Economic costs, environmental benefits and cost-effectiveness

Figure 3-5 shows information communications of economic costs, environmental

benefits, and cost-effectiveness of BMPs. An integrated economic-hydrologic model can be the

engine to provide the information content. A WebGIS interface can be used to support the

examination of “What if” BMP scenarios, the presentation of modelling results, and the

interaction with the modelling results by both farmers and conservation managers (Shao et al.

2017).

Web pages

(i.e. Agri-environmental policy and BMP related technical

knowledge)

BMP technical documentation

(i.e. BMP related technical knowledge)

Conservation managers

Information portal

Farmers Conservation managers

are collected by

import information into

is explored by

46

BMP policy/management

BMP policy/management at a watershed scale has two tasks: minimizing BMP costs

subject to environmental targets and maximizing environmental effectives subject to financial

constraints (Shao et al. 2017). As shown in Figure 3-5, an optimization model can be used to

fulfil the two tasks. A WebGIS interface can be used to present the BMP policy/management

information to conservation managers. Conservation managers can use the information to

allocate BMP investment within the watershed and incentivize farmers to adopt BMPs at optimal

locations.

Figure 3-5: The information sub-model for BMP planning

Integrated hydrologic-economic modelling An optimization model

Economic costs, environmental benefits, and cost-effectiveness

of BMPsBMP policy/management

WebGIS

Farmers Conservation managers

Discussion forum & E-mail

produce produce

asynchronously communicate using

is explored by

is presented on

47

Communications between farmers and conservation managers

The communications between farmers and conservation managers are essential for

addressing various questions on BMP costs, benefits, and cost-effectiveness and build a mutual

consensus on BMP adoption. The WebGIS allows more flexible, asynchronous communications

between farmers and conservation managers. The asynchronous communications can be

supported by email and discussion forum. When one person sends a message to another person,

the message receiver is notified and can reply to the message sender when available. Meanwhile,

the communication history can be maintained for both the sender and receiver to review and

track their communications.

3.3 Summary

Agricultural BMP adoption can be viewed as a process wherein stakeholders including

farmers and conservation managers achieve a consensus on BMP implementation through

information communications. Information needs of farmers and conservation managers can be

classified into two categories: public information and BMP planning information. Based on their

information needs, an information model is developed to characterize the information

communication process for BMP adoption, which involves information contents,

communications, and related technologies and tools. The information model includes three sub-

models. Specifically, one information sub-model is developed to characterize public information

communications on field characteristics, environmental concerns, and BMP adoption, one

information sub-model is developed to characterize public information communications on agri-

environmental policy and BMP related technical knowledge, and one information sub-model is

developed to characterize confidential information communications on BMP planning, which

48

including information on BMP costs, effectiveness, and cost-effectiveness for supporting the

examination of "What if" BMP scenarios by farmers and also BMP policy/management tasks by

conservation managers.

49

Chapter 4 System Architecture and Design of the WebGIS-based Decision Support System for Facilitating Agricultural BMP

Adoption

This chapter presents the architecture and design of the WebGIS-based decision support

system for facilitating agricultural BMP adoption. The design of the WebGIS-based decision

support system implements the information model to support the communication process of

public and BMP planning information in agricultural BMP adoption. The content of this chapter

is organized into five sections. The first section introduces a task-oriented design approach to

design the WebGIS-based decision support system. A hierarchical task analysis is used to guide

the design of the system at three design levels: the subsystem, module and component levels.

The second section introduces the loose-coupling between the WebGIS and the watershed

modelling tools, including an integrated economic-hydrologic model and an optimization model.

The third section introduces the subsystem of the WebGIS-based decision support system, which

includes the public subsystem, the BMP planning subsystem, and the administration subsystem.

In the fourth section, the modules of the three subsystems are given, and the fifth section

introduces the components of each module.

4.1 Task-oriented design of the WebGIS-based decision support system for facilitating agricultural BMP adoption

As an extension to the traditional desktop-based GIS system for BMP evaluation, the

WebGIS-based decision support system for BMP adoption needs to meet a wide range of system

requirements. In particular, considering that the WebGIS-based decision support system is

intended to support farmers and conservation managers instead of modelling experts and

researchers, the system has to be easy to use for novice users. Many literatures from Human

50

Computer Interaction (HCI) have described what features an “easy-to-use” system. For example,

Fischer, G. (1993) and Carroll, J.M. (1997) suggested that an “easy-to-use” system should be

user-centered and meet users’ task requirements. The understanding of user tasks is essential to

develop system interactions to meet the task requirements (Fischer, G., 2001).

Thus, a task-oriented design approach is used to design the WebGIS-based decision

support system for facilitating agricultural BMP adoption. Different from the traditional

“Waterfall” model of system design which focuses on system functions, the task-oriented design

approach focuses on representative user tasks, and emphasizes the design of an effective, user-

friendly interface (Lewis & Rieman, 1994). The task-oriented design facilitates the

implementation of the information model in two steps. Firstly, based on a hierarchical task

analysis, a hierarchy of user tasks is developed to understand how the information model is

implemented by user tasks or what user tasks in the information communication process are for

BMP adoption. Secondly, the hierarchy of user tasks is applied to guide the design of the

WebGIS-based decision support system at three system design levels: the subsystem, module,

and component levels.

4.1.1 A hierarchy of user tasks

The task-oriented design approach requires an in-depth understanding on the user tasks

for agricultural BMP adoption. In this regard, this research uses a hierarchical task analysis to

understand the structure of user tasks in the information communication process for BMP

adoption. According to Annett, J., (2004), the hierarchical task analysis provides an effective

framework to facilitate the structuring of user tasks as well as specifying the task-related

constraints, such as location, time, tools and conditions for completing the tasks.

51

The hierarchy of user tasks includes three levels. The task level one corresponds to the

information needs: the public and BMP planning information. It addresses general questions

such as “Bob would like to learn more about the BMP concept and implementation” and “Bob

would like to use the BMP planning information to examine BMP scenarios on his field”. The

task level two includes information tasks required to achieve the information needs. Two types

of information tasks at this level are mainly related to information provision and communication.

Sample tasks can include “Bob would like to evaluate the BMP cost-effectiveness” and “Bob

would like to discuss BMP adoption with conservation managers”. These two tasks are subtasks

to support Bob to plan BMPs on his fields which result from refining the tasks defined at level

one. Lastly, the task level three focuses on user interactions with the system. It defines how the

users interact with the system to conduct the information tasks defined at level two (i.e.

information provision and communication) and how the components are designed to support user

interactions. An example of tasks at this level can be “Bob opens an online discussion window to

initiate a discussion on BMP planning”. The tasks at this level support the specifications of the

system interface design.

To make sure the tasks are realistic, complete and representative, the development of the

task hierarchy used two methods. Firstly, the tasks were discussed and verified with

representative users from conservation authority and government. In meetings and workshops,

the workflow of tasks was demonstrated. Feedback from the representative users was collected to

refine the tasks. Secondly, the tasks were referenced and further refined based on a desktop GIS-

based system for watershed evaluation of BMPs (Shao et al., 2017). The GIS system

characterized the agricultural BMP planning process and was evaluated by both representative

users and experts.

52

4.1.2 System design levels

According to the hierarchy of user tasks, the WebGIS-based decision support system is

designed at three levels: the subsystem level, the module level, and the component level. Each of

these levels corresponds to a level of tasks (Figure 4-1). Specifically, the subsystems aim to

support the information needs (i.e. the public and BMP planning information), the module aims

to support information tasks (i.e. information provision and communication) to meet the

information needs, and the components aims to support user interactive operations to conduct

information tasks.

Figure 4-1: Task-oriented design of the WebGIS-based decision support system for facilitating agricultural BMP adoption

Subsystems

Modules

Components

System design levels Hierarchy of user tasks

Task Level 1Information needs

Task Level 2Information tasks

(Information provision and communication)

Task Level 3Interactive operations

achieve

support

decompose

decompose

achieve

support

support

53

4.2 Integrating WebGIS and watershed modelling tools

Watershed modelling tools generate agri-environmental planning information on

economic costs, environmental benefits and cost effectiveness of BMPs. Integrating WebGIS

and watershed modelling tools is necessary for farmers and conservation managers to easily

access the information and conduct BMP planning tasks. In this study, the WebGIS and the

watershed modelling tools are integrated using a loose-coupling approach. Two watershed

modelling tools include an integrated economic-hydrologic model and an optimization model

(Shao et al., 2017). In the loose-coupling approach, the GIS and watershed modelling tools

communicate through external files. To increase the efficiency of coupling, the communication

process between the GIS and watershed modelling tools are automated: program routines are

designed to automatically generate modelling input files, process modelling output files, and

present information on the WebGIS.

4.2.1 WebGIS and the integrated economic-hydrologic model

The loose coupling of WebGIS and the integrated economic-hydrologic model is shown

in Figure 4-2. The WebGIS supports the preparation of user-specific modelling inputs, including

farm fields and BMP assignments on these fields. The SWAT model uses the user-specific

modelling inputs combined with pre-defined modelling inputs such as land use data and

management data to generate information on environmental effects of “What if” BMP scenario,

such as sediment, total phosphorus, and flow. The economic model generates information on the

economic effects of BMPs, including revenue, cost and net return.

54

Figure 4-2: Loose coupling of WebGIS and the integrated economic-hydrologic model

An integrated economic-environmental analysis of BMP scenarios is based on the

combination of modelling results from both the SWAT models the economic model. Because the

SWAT model simulates the hydrologic process based on hydrologic response units (HRU), the

hydrologic modelling results are interpolated into the field scale. The combined results are stored

in spatial databases and GIS files. These data sources can then be used by WebGIS for result

presentation. The spatial databases can be used to further process information, draw charts and

generate result tables. The GIS files can be used by the WebGIS to render interactive modelling

result maps.

4.2.2 WebGIS and the optimization model

Coupling WebGIS and the optimization model is shown in Figure 4-3. The WebGIS is

used to prepare user-specific modelling input for the optimization model, including the farm

fields, BMP types, and optimization objectives.

The SWAT model The economic model

Processing

WebGIS

Spatial databases and GIS files

55

Figure 4-3: Loose coupling of the WebGIS and the optimization model

Based on user-specific inputs, the optimization model generates the BMP

policy/management information at a watershed scale. The BMP policy/management information

specifies how to plan BMPs on the targeted fields to achieve environmental or economic

objectives to achieve cost-effectiveness. An open-source Mixed Integer Linear Programming

Solver is embedded in the model to identify field-specific BMP combinations with cost

minimization objective function subject to environment targets or with maximizing

environmental benefit objective function subject to economic constraints (Oginskyy, 2014). A

field-BMP combination can be any combination of field and BMPs, such as field 1 with cover

crop + conservation tillage or field 2 with nutrient management. The solver works based on

ranking the cost-effectiveness of all the field-BMP combinations. The field-BMP combinations

with the highest cost-effectiveness are considered as the most cost-effective. To support the

optimization modelling, a database is created to store data on the economic and environmental

effects of BMP combinations on each farm field based on on-farm economic modelling and

watershed hydrologic modelling.

The optimization model

WebGIS

Spatial databases and GIS files

Optimization databases

56

4.3 The subsystems of the WebGIS-based decision support system for facilitating agricultural BMP adoption

The WebGIS-based decision support system is designed to include three subsystems: a

public subsystem, a BMP planning subsystem and an administration subsystem (Figure 4-4).

The public subsystem supports the communication process of public information. The

user groups of the public subsystem include farmers and conservation managers. It aims to

support tasks such as information sharing among farmers about their field characteristics,

environmental concerns, and BMP adoption, and dissemination of agri-environmental policies

and BMP related technical knowledge. The public subsystem also supports BMP related

communications among farmers.

The BMP planning subsystem supports the communication process of BMP planning

information, which can be used by only registered farmers and conservation managers due to

confidentiality of the BMP planning information. The subsystem supports farmers and

conservation managers to evaluate “What if” BMP scenarios and answer questions like “What

are the environmental and economic effects of BMPs if the BMPs are applied to specific farm

fields”. The subsystem also supports conservation mangers to answer two policy/management

design questions: “How to implement BMPs in targeted fields to meet an environmental target

with minimized costs” and “How to implement BMPs in targeted fields to maximize

environmental benefits under a financial constraint”. The subsystem also supports discussions

between farmers and conservation managers to address various BMP adoption questions.

Different from the public and BMP planning subsystems which support the information

communication process, the administration subsystem supports system operations. It allows the

57

administrator to register users to access the BMP planning subsystem. It also supports

administrators to monitor the use and collect usage information of BMP planning subsystem by

farmers and conservation managers. Such usage information could be used to evaluate system

activities of farmers and conservation managers and understand their progress on various BMP

assessment tasks.

58

Figure 4-4: Subsystems in the WebGIS-based decision support system for facilitating agricultural BMP adoption

WebGIS-based decision support system for facilitating agricultural BMP adoption

The public subsystem The BMP planning subsystem The administration subsystem

• Economic costs, environmental benefits and cost-effectiveness of BMPs

• BMP policy subject to environmental targets

or economic constraints

• Communications between farmers and

conservation managers

• Field characteristics, environmental

concerns, and BMP adoption

• Agri-environmental policies

• BMP related technical knowledge

• Communications among farmers

• Usage information of the BMP planning subsystem

59

4.4 The modules of the three subsystems

This section describes the modules within the three subsystems of the WebGIS-based

decision support system: the public subsystem, the BMP planning subsystem, and the

administration subsystem. The modules aim to support information tasks in the subsystems.

4.4.1 The modules of the public subsystem

The public subsystem facilitates the communication process of the public information.

The subsystem comprises of two modules: “Information sharing site” and “Public information

center” (Table 4-1).

Table 4-1: Tasks of system modules in the public subsystem

Module Tasks

Information sharing site Submit and view public annotations

Public information center Search for information and explore searching results

Information sharing site: The main objective of the “Information sharing site” module is

to support information sharing among farmers about their field characteristics (e.g. crop type,

soil type, and soil quality), environmental concerns (e.g. erosion), and BMP adoption (e.g.

conservation tillage). The module also supports farmers to develop communications on BMP

related topics. To support these objectives, the information sharing site enables farmers to submit

public annotations and view the public annotations submitted by other farmers.

60

Public information center: the “Public information center” is designed to disseminate

public information such as agri-environmental policies and BMP related technical knowledge.

The module allows farmers to search and explore the information of their interest.

4.4.2 The modules of the BMP planning subsystem

The BMP planning subsystem is designed to support the communication process of BMP

planning information. The subsystem is composed of five modules including “Access control”,

“Scenario exploration”, “Policy/Management”, “Discussion” and “Report”. These modules

support several information tasks; some complex tasks can be decomposed into several steps

(Table 4-2).

Access control: The “Access control” module defines how the system access should be

granted to farmers and conservation managers to use the BMP planning subsystem. It allows

only the authorized farmers and conservation managers to log into the BMP planning subsystem.

Scenario exploration: The “Scenario exploration” module supports farmers and

conservation managers to explore “What if” BMP scenarios to address questions like what the

BMP economic costs, environmental benefits and cost-effectiveness are if BMPs are

implemented in specific farm fields. To facilitate this task, the module handles several task steps,

including scenario creation, scenario development, scenario evaluation, and scenario

comparison.

Policy/Management: The “Policy/Management” module supports conservation managers

to design two types of BMP policies including spatial targeting of BMPs to achieve

environmental targets with minimized costs and maximizing environmental benefits subject to

61

financial constraints. To support the policy design, the module has two components including

scenario optimization and optimization result exploration.

Table 4-2: Tasks of system modules in the BMP planning subsystem

Module Task Task steps

Access control Authorize farmers and conservation

managers to use the BMP planning

subsystem

N/A

Scenario exploration Evaluate economic costs, environmental

benefits, and cost-effectiveness of BMPs

Create a BMP scenario

Develop a BMP scenario

Evaluate a BMP scenario

Compare the BMP scenario with a

baseline scenario

Policy/Management Design BMP policy/management subject to

economic and environmental constraints

Optimize a BMP scenario

Explore the optimization results

Discussion Submit discussion topics and reply N/A

Report Generate scenario reports in HTML/PDF

format

N/A

Discussion: The “Discussion” module enables farmers and conservation managers to

communicate on various BMP related topics. Farmers and conservation managers can create

discussion topics, and submit, view and reply discussion comments.

Report: The “Report” module facilitates the communication process by automatically

generating reports for the “Scenario exploration” and “Policy/Management” results. The module

can generate both HTML and PDF reports.

Table 4-2 summarizes the tasks and steps in the modules of the BMP planning

subsystems. “Scenario exploration” and “Policy/Management” modules are the key modules to

support BMP planning. The tasks of these two modules are analyzed using the scenario scripts or

62

narratives (Appendix A). The scenario scripts for eliciting tasks in the two modules are generated

based on the GIS-based system for watershed evaluation of BMPs (Shao et al., 2017) and also

inputs from conservation authority and government.

4.4.3 The modules of the administration subsystem

The administration subsystem supports user registration and system monitoring activities.

The administration subsystem includes three modules (Table 4-3).

Table 4-3: Tasks of system modules in the administration subsystem

Module Tasks

Access control Authorize the system administrator to use the administration subsystem

User registration Register users to use the BMP planning subsystem

System monitoring Analyze and display information on user activities within the BMP

planning subsystem

Access control: The “Access control” module is designed to authorize the system

administrator to use the administration subsystem.

User registration: The “User registration” module is for the system administrator to

register new users (i.e. farmers or conservation mangers) to use the BMP planning subsystem.

System monitoring: The “System monitoring” module is designed to support the

administrator to monitor the activities of farmers and conservation managers in the BMP

planning subsystem and understand the progress of their BMP assessment tasks.

63

4.5 The components of the system modules

This section discusses the design of modules in the WebGIS-based decision support

system. Each module has three tiers including client, service and data tiers. In the client-server

architecture, the client tier is on the client side and the service and data tiers are on the server

side. The client tier is designed to present information and support user interactions, the service

tier is designed to handle information requests from client and generates information required for

client presentation, and the data tier is designed to manage module-required data and files.

Components of the three tiers within each module are introduced in this section.

Specifically, the components on the client tier (i.e. user interface elements) are identified, the

components on the service tier (i.e. web services) are explained, and the components on the data

tier (i.e. database tables) are illustrated. The communication processes among these components

to address user information requests are also explained.

4.5.1 The components of the modules in the public subsystem

4.5.1.1 Module: Information sharing site

The “Information sharing site” module supports farmers to share information on their

field characteristics, environmental concerns, and BMP adoption. The component diagram of the

module is shown in Figure 4-5. To realize the module, three client components are designed: an

annotation form is designed for farmers to input information for sharing, an interactive map is

designed to present the submitted annotations, and an annotation list is designed to organize

annotations by categories and support filtering annotations by annotation contents.

64

To support farmers to submit and view annotations, several interactions are designed by

attaching user actions to the client components. Firstly, a click event is attached to the interactive

map. When a farmer clicks on the map, an annotation form will be triggered for the farmer to

input the information. Secondly, a click event is attached to the annotation form. When a farmer

submits the form, a request will be sent to the service tier for annotation submission. Thirdly, all

the three client components are interlinked with each other. When an annotation is submitted

from the annotation form, both interactive map and annotation list will be updated: an annotation

marker will be added to the interactive map, and an annotation will be inserted into the

annotation list. Fourthly, a filter is implemented on the annotation list, which can be used to filter

annotations by their information contents or categories. Finally, a “hover” map event is attached

to the interactive map. When a user hovers on the annotation marker, the annotation marker on

the map and the corresponding annotation on the annotation list will be highlighted.

Figure 4-5: Component diagram of the “Information sharing site” module

Two web services are designed in the service tier to support annotation submission and

loading/viewing. The “Write public annotation” web service writes submitted annotations to the

Interactive map

Write public annotation

Public annotation table

Annotation form Annotation list

Read public annotation

Render Render

ReadWrite

Update Update

Sent

Client

Service

Data

Trigger Filter

65

public annotation table, and the “Read public annotation” web service loads annotations from the

table and sends them to the map and annotation list for presentation. To support the annotation

management, a public annotation table is designed (see Appendix B, Figure B.1).

4.5.1.2 Module: Public information center

The “Public information center” module aims to facilitate farmers to obtain information

on agri-environmental policies and BMP related technical knowledge of their interest. The

component diagram of the module is shown in Figure 4-6. Two client components are designed

in the client tier including a searching tool and an information table. The searching tool allows

farmers to input a keyword to search for the public information. The information table is used to

display the searching results. A “click” event is registered to the searching tool; when the

keyword is submitted, a searching request will be sent to the service tier for obtaining results.

A web service called “Public information management” is designed to support

information searching. Upon receiving the searching request, the web service will read the three

public information tables (i.e. News table, Policy table and “BMP technical knowledge” table)

and fetch the matched information with the searching keyword. The three tables are created to

maintain different categories of public information (see Appendix B, Figure B.2).

66

Figure 4-6: Component diagram of the “Public information center” module

4.5.2 The components of the modules in the BMP planning subsystem

4.5.2.1 Module: Access control

The “Access control” module allows authorized farmers and conservation managers to

enter the BMP planning system. The component diagram of the module is illustrated as Figure 4-

7. In the client tier, a login form is designed to allow a user to input his/her name and password

to enter the subsystem. A “click” event is registered to the form to send a request to the service

tier for user authorization.

In the service tier, the “User authorization” web service is designed to handle the login

request. The web service is implemented in two steps. Firstly, the web service reads the user

table (see Appendix B, Figure B.3) and finds a user record that matches the user name and

password sent from the client to identify the role of the user. Secondly, based on role of the user,

the web service determines the subsystem that the user is allowed to enter. If the user is a farmer

or a conservation manager, the web service will direct him/her to the BMP planning subsystem.

Searching tool

News table

Information table

Public information management

Render

Read

Sent

Client

Service

Data Policy table “BMP technical knowledge” table

Read Read

67

If the user is an administrator, the web service directs the user to the administration subsystem.

This module is shared by both the BMP planning subsystem and the administration subsystem.

Figure 4-7: Component diagram of the “Access control” module

After authorization, the web service will send the user information and direct the user to

the BMP planning subsystem. The user information will be used to control the visibility of fields

on the interactive map within the BMP planning subsystem. In the BMP planning subsystem,

farmers can only view their own farm fields and assign BMPs on their own fields, while

conservation managers can assign BMPs on all the fields in the watershed.

4.5.2.2 Module: Scenario exploration

The “Scenario exploration” module aims to support farmers and conservation managers

to explore “What if” BMP scenarios to understand the economic costs, environmental benefits,

and cost-effectiveness of BMPs. The design of the module requires fulfillment of several task

steps, including BMP scenario creation, BMP scenario development, BMP scenario evaluation,

and BMP scenario comparison.

Login form

User Table

User authorization

Read

Sent

Client

Service

Data

The BMP planning system

Conservation manager and farmer

Administrator

The administration

system

Role

68

Task step: BMP scenario creation

Figure 4-8 shows the component diagram for BMP scenario creation. To support the task,

a scenario creation form is designed. The required information for creating a scenario includes a

scenario name and a scenario description. A “click” event is attached to the scenario creation

form to send information to the service tier for scenario creation.

The “Scenario creation” web service is designed to handle the BMP scenario creation

request. The service writes the scenario information into a scenario table (see Appendix B, Table

B.4). It also updates the system interface using a HTML template for BMP scenario

development, which is the next step.

Figure 4-8: Component diagram for BMP scenario creation

Task step: BMP scenario development

The component diagram for BMP scenario development is shown in Figure 4-9. To

support the scenario development, three components are designed in the client tier: a button

group is used to select BMP types, an interactive map is used to display different BMP layers,

and a BMP assignment table is used to store the status of BMP assignments. The interactive map

Scenario creation form

Scenario table

Scenario creation

Write

Sent

Client

Service

Data

RenderHTML template

for scenario development

69

is composed of several BMP layers and only one BMP layer is active or visible based on user

selection. The visibility of BMP layer is controlled by the button group. For example, when the

BMP type “conservation tillage” is selected, the map will only show the conservation tillage map

layer for the user to assign conservation tillage to farm fields. When the BMP is assigned to

fields, the BMP assignment table will be updated to reflect the current status of BMP

assignments.

The “Scenario development” web service is triggered when the BMP assignment table is

sent to the service tier. It writes the BMP assignment information into a BMP configuration table

(see Appendix B, Figure B.5). Meanwhile, it renders a webpage using a HTML template for

BMP scenario evaluation, which is the next step.

Figure 4-9: Component diagram for BMP scenario development

Task step: BMP scenario evaluation

The component diagram for BMP scenario evaluation is as indicated in Figure 4-10.

When the “Scenario development” web service renders the HTML template for scenario

evaluation, a request for BMP scenario evaluation is sent. The “Scenario evaluation” web service

Interactive map

BMP configuration table

Scenario development

Write

Send

Client

Service

Data

BMP assignment table

UpdateButton group

Control

HTML template for scenario evaluation

Render

70

is designed to handle the scenario evaluation request. The request is handled in two steps. Firstly,

the “Scenario evaluation” web service calls the integrated economic-hydrologic model to

generate the scenario evaluation results, i.e. economic costs and environmental benefits of the

BMP scenario. The scenario evaluation results are written into a modelling result table (see

Appendix B, Table B.6). Secondly, the “Scenario evaluation” web service reads the modelling

results and generates the scenario evaluation map file.

To support the presentation of the scenario evaluation results, two components are

designed in the client tier. An interactive map is used to render the scenario evaluation map file,

and an evaluation chart is used to visualize the BMP evaluation results within a time frame. The

data for rendering the evaluation chart is read from the modelling result table and sent to the

client by the “Scenario evaluation” web service.

Figure 4-10: Component diagram for BMP scenario evaluation

Figure 4-11 shows the component diagram for exploring the evaluation results. In the

client tier, a button group is used for a user to examine the modelling results by different

variables (e.g. total phosphorous, total nitrogen, sediment, cost and revenue). Clicking a button

Scenario evaluation map file

Scenario evaluation

Write

Sent

Client

Service

Data

BMP assignment table

Modelling result table

Write

Interactive map

Read

Evaluation chart

Render

Integrated economic-hydrologic model

Call

Read Write

71

in the button group will trigger two actions. Firstly, the interactive map will be updated to show

the results of the selected variable. Secondly, the “Draw evaluation chart” web service will be

triggered to read the results of the selected variable from the modelling result table and update

the evaluation chart. For example, when the “sediment” button is clicked, the modelling results

of sediment will be rendered on the map and the evaluation chart will be updated accordingly.

Figure 4-11: Component diagram for BMP evaluation result exploration

Task step: BMP scenario comparison

After the completion of BMP scenario evaluation, the subsystem allows the user to

compare the BMP scenario with a baseline scenario. The component diagram for scenario

comparison is shown in Figure 4-12. To support scenario comparison, three components in the

client tier are designed. A scenario list is used for users to select a baseline scenario to compare

with, an interactive map is used to display the comparison results, and a button group is used for

users to investigate the scenario comparison results for different variables (e.g. total

phosphorous, total nitrogen, sediment, cost and revenue).

The scenario comparison starts when a user selects a scenario on the scenario list and

sends a scenario comparison request to the service tier. The “Scenario comparison” web service

Draw evaluation chart

Sent

Client

Service

Data

Button group

Modelling result table

Read

Interactive mapUpdate

Evaluation chart

Update

72

is designed to handle the request. To make the comparison, it reads the results of the baseline

scenario, calculates the difference between the two scenarios, and writes the comparison results

into a comparison map file. When the map file is generated, the file will be read by the

interactive map for presentation. The comparison between the “What if” BMP scenario and the

baseline scenario will show the effect of BMPs including BMP costs, benefits, and cost-

effectiveness.

Figure 4-12: Component diagram for BMP scenario comparison

4.5.2.3 Module: Policy/Management

The “Policy/Management” module aims to supports conservation managers to design

BMP policy at a watershed scale. The module achieves this objective in two steps. Firstly, the

module allows scenario optimization to generate the BMP policy/management information.

Secondly, the module provides tools for users to explore the optimization results.

Scenario comparison

Sent

Client

Service

Data

Scenario list

Modelling result table

Read

Interactive map

Comparison map file

Render

Write

Button group

Update

73

Task step: Scenario optimization

Figure 4-13 shows the component diagram for scenario optimization. In the client tier, an

optimization input component is designed for users to submit required parameters for BMP

policy/management design, which include the policy/management type (economic vs.

environmental), the pollutant type, a set of selected farm fields and an optimization constraint

(environmental vs, economic). In addition, an interactive map, an optimization chart, and an

optimization table are designed for the optimization result presentation. The optimization result

shows the configurations of BMPs in farm fields to achieve specific environmental or economic

objectives. The interactive map allows users to examine the BMP configuration in farm fields,

the optimization chart allows users to examine the trade-offs between the economic costs and

environmental benefits, and the optimization table allows users to examine the BMP

configuration by farm field through a table list.

Figure 4-13: Component diagram for scenario optimization

To generate the BMP policy/management information, the parameters for

policy/management design are sent to the server. The optimization request is handled by the

“Scenario optimization” web service in two steps. Firstly, it calls an optimization model to

Optimization result map file

Scenario optimization

Write

Read

Client

Service

Data

Optimization input

Optimization result table

Interactive map

Write

Optimization chart

Sent

Optimization table

An optimization modelCall

Render

Read

74

generate the optimization results and write the results into an optimization result table (see

Appendix B, Figure B.7). Secondly, it reads the optimization results from the table and generates

a map file and at the same time, sends the optimization results to the client to render the

optimization chart and optimization table. This generated map file will be read by the interactive

map to present the policy/management information.

Task step: Exploration of the optimization results

The generated BMP policy/management information can be explored through the

optimization chart. Figure 4-14 shows the component diagram for the information exploration. A

“click” event is attached to several key points on the optimization chart to allow users to

investigate the trade-off between the economic costs and environmental benefits of BMP

configurations in fields. Each key point represents a configuration of BMPs with the

corresponding economic costs and environmental benefits at that point. When the “click” event

is triggered and a key point is selected, the “Draw optimization chart” web service will be

triggered. The service will read the optimization results from the optimization result table and

send it back to the client. The retrieved information will be used to render the interactive map

and the optimization table to reflect the BMP configuration at the selected key point on the chart.

Figure 4-14: Component diagram for optimization result exploration

Draw optimization chart

Read

Client

Service

Data Optimization result table

Interactive map Optimization chart Optimization table

Sent

Render

Button groupControl

Update

75

4.5.2.4 Module: Discussion

The “Discussion” module aims to support two tasks. One is to submit discussion topics.

Another is to submit discussion replies. Figure 4-15 shows the component diagram of the

“Discussion” module. Three components are designed in the client tier: a discussion topic form

is used to support discussion topic submission, a discussion reply form is used to support

discussion reply submission, and a discussion window is used to show discussion threads which

organize discussion topics and replies. To enable user interactions with the components, a

“submission” event is attached to both the discussion topic form and discussion reply form.

When a form is submitted, the request will be sent to the service tier.

Two web services are designed to support the two tasks. The “Discussion topic creation”

web service is designed to write discussion topics into a discussion topic table (see Appendix B,

Figure B.8) and update the topic in the discussion window. The “Discussion thread

management” web service is designed to write discussion replies into the discussion thread table,

update the discussion network table, and update the discussion thread in the discussion window.

The discussion thread table is used to maintain discussion replies under a discussion topic (see

Appendix B, Table B.9). The discussion network table is used to maintain the number of

communications between two users (see Appendix B, Table B.10).

76

Figure 4-15: Component diagram of the “Discussion” module

4.5.2.5 Module: Report

The “Report” module supports farmers and conservation managers to generate HTML

and PDF reports from the “Scenario exploration” and “Policy/Management” results. The

component diagram of the “Report” module is shown in Figure 4-16. The client tier of the

module is composed of a button group which includes two separated buttons. One button is for

HTML report generation and another is for PDF report generation. Both buttons support a

“click” event to send a report generation request to the service tier for report generation.

A “Report generator” web service is designed to handle the report generation request.

The logic of the service is designed in two steps. Firstly, it gathers information for report

generation, which includes BMP configuration information from the BMP configuration table,

and scenario evaluation, comparison, and optimization results. Secondly, the web service

chooses different methods to generate the report. To generate the HTML report, the web service

inserts the collected information into a HTML report template. To generate the PDF report, the

Discussion topic form

Discussion topic table

Discussion topic creation

Write

Submit

Client

Service

Data Discussion thread table

Discussion thread management

Discussion reply form

Submit

Discussion network table

Write Update

Discussion window

Update Update

77

service inserts the collected information into a text report template and then converts the

template into a PDF using the Pandoc – a software to convert files between different formats.

Figure 4-16: Component diagram of the “Report” module

4.5.3 The components of the modules in the administration subsystem

4.5.3.1 Module: User registration

The module of “User registration” supports the system administrator to register new users

to access the BMP planning subsystem. The component diagram of the module is shown in

Figure 4-17. A user registration form is designed in the client tier to allow the administrator to

enter necessary information for the registration, which includes the user’s name, email address,

the role of the user and the user’s farm number. The role of a user can be either farmer or

conservation manager. The user’s farm number is used to link the user with his/her farm fields.

This linkage is important as it controls the availability/visibility of farm fields when a user logs

into the BMP planning subsystem. Farmers can view only their own farm fields, while the

conservation manager can view all the fields in a watershed.

Button group

Report generator

Read

Client

Service

Data BMP configuration table

HTML report template

Insert

Send

CallPandoc

HTMLor

PDF

78

Figure 4-17: Component diagram of the “User registration” module

The registration process starts with the registration request sent from the client. A “User

registration” web service handles the request. Upon receiving the user information for

registration, the service will write the user information into the user table (see Appendix B, Table

B.3) and at the same time, prepare modelling files for the user. Those modelling files contain the

integrated economic-hydrologic models and the optimization model for running the modules of

“Scenario exploration” and “Policy/Management”. Once the user registration completes, the

service will send an email to the registered email address to notify the success of registration.

Because the email client is not within the scope of this module design, the component of email

client in the diagram is within a dotted rectangle.

4.5.3.2 Module: System monitoring

The “System monitoring” module provides monitoring information on user activities in

the BMP planning subsystem. The component diagram of the module is shown in Figure 4-18.

To support monitoring tasks, three components are designed in the client tier: two tables to

present scenario information, and a communication network to present the frequency of

User registration form

User Table

User registration

PrepareWrite

Sent

Client

Service

Data Modelling files

Send emailEmail client

79

communications among users. A summary information table is used to present summarized

monitoring information on the usage of the BMP planning subsystem, such as the total number

of scenarios created and completed, and total number of discussion topics and replies submitted.

A personal information table is used to present personal monitoring information, such as the total

number of scenarios created and completed, and the total number of discussion topics and replies

submitted by each user.

The communication network is designed to illustrate the frequency of communications

among users. It is designed as a bi-directed graph. Each node on the graph represents a user (a

farmer or a conservation manager); each connection on the graph that connects two nodes

represents a communication link between the two users. The activeness of a user for

communications is shown by the size of the node – more active users are presented as bigger

nodes.

Figure 4-18: Component diagram of the “System monitoring” module

The information in the tables and the communication network is obtained by two web

services. The “Scenario management” web service is designed to obtain information to render

Summary information table

Scenario table

Scenario management

Read

Client

Service

Data Discussion topic table Discussion reply table

Render

Personal information table Community network

Communication management

Read

Discussion network table

Render

80

the two tables. The information is generated by joining three tables: the scenario table, the

discussion topic table, and the discussion reply table. The “Communication management” web

service is designed to render the communication network from the discussion network table.

4.6 Summary

This chapter presents the system architecture and design of the WebGIS-based decision

support system for facilitating agricultural BMP adoption. To implement the information model

for BMP adoption, the WebGIS-based decision support system is designed into three subsystems

including a public subsystem, a BMP planning subsystem and an administration subsystem.

The public subsystem includes 1) an “Information sharing site” module to share

information on their field characteristics, environmental concerns, and BMP adoption using

annotations and 2) a “Public information center” module to facilitate farmers to obtain

information on agri-environmental policies and BMP related technical knowledge of their

interest using database search functions.

The BMP planning subsystem includes 1) an “Access control” module for restricting the

users to farmers and/or conservation managers, 2) a “Scenario exploration” module with “What

if” BMP scenario creation, scenario development, scenario evaluation, and scenario comparison

functions, 3) a “Policy/Management” module for spatial targeting of BMPs based on either

environmental targets or financial constraints, 4) a “Discussion” module for submitting and

replying to various BMP related topics, and 5) a “Report” module for generating BMP

evaluation and/or policy/management reports in HTML or PDF formats.

81

The administration subsystem includes 1) a “User registration” module that supports the

system administrator to register new users to access the BMP planning subsystem and 2) a

“System monitoring” module that provides information on user activities in the system.

82

Chapter 5 A Prototype of the WebGIS-based Decision Support System for the Gully Creek Watershed

This chapter presents a prototype of the WebGIS-based decision support system for a

representative study watershed. The chapter starts with the introduction of the study area, the

Gully Creek watershed. In the second section, the development of the system prototype is

introduced: the software for the development is introduced, system localization for the Gully

Creek watershed is discussed, and a demonstration of the system prototype is given to illustrate

the functionalities of the WebGIS-based decision support system prototype to support

agricultural BMP adoption in the study area.

5.1 Study area

The 14.3-km2 Gully Creek watershed is a representative watershed of a series of small

watersheds along the shoreline of Lake Huron (Figure 5-1). Similar to other lakeshore streams,

Gully Creek discharges directly into Lake Huron. Due to its potential to influence near shore

water quality, the watershed has been classified as an Environmentally Sensitive Area (Veliz et

al., 2006). The Gully Creek watershed has an undulating terrain, typical of the small lakeshore

watersheds. Land elevations of the watershed range from 176 to 281 m. The average slope in the

watershed is 6% with a minimum of 0% in flat areas and as high as 95% in incised gully areas

(typically greater than 9% in riparian areas). About 70% of the land is agricultural and 25% is

natural vegetation, including trees, shrubs and grasses. This natural vegetation primarily buffers

the main channel. Corn, soybean and winter wheat are the main crops grown in the watershed.

With growing concerns about near-shore water quality of Lake Huron, BMP implementation in

shoreline watersheds has become one of the important measures for mitigating these negative

83

effects. Representative BMPs in the Gully Creek watershed include conservation tillage, nutrient

management, cover crop, and water and sediment control basins (WASCoBs).

Figure 5-1: The Gully Creek Watershed in Southern Ontario, Canada (Shao et al., 2017)

84

5.2 The prototype development for the Gully Creek watershed

5.2.1 Principles for software selection

Based on the design of the WebGIS-based decision support system, the implementation

involves technologies and tools for both client (i.e. client tier) and server (i.e. service and data

tier) development. This section will introduce the principles for software selection to implement

the WebGIS-based decision support system. Key software technologies and their roles for the

implementation are discussed.

The system development utilizes several server-side and client-side software products.

Given that various technological choices are available on the current market for system

development, a careful software selection process is necessary. For this development, the

selection of the software products follows three main principles.

Firstly, the system development uses mainstream software products with better technical

support and broader user communities. When deciding between two competing software

products, the development chooses the product that is intensively evaluated and stably updated.

For example, for the client interface development, a gridding framework is necessary for User

Interface component alignment on the browser. Both Bootstrap and Foundation are popular

gridding frameworks on the current market. Bootstrap is selected for this development as it

provides more themes and has a bigger user community (“Bootstrap vs. Foundation”, 2019).

Secondly, the development and maintenance costs for the software should be low. For

this reason, free and open source software is favored for the development. For example, while

many online mapping software exist, OpenLayers is selected for developing online mapping

85

functionalities (Gratier et al., 2015). Google Map is not chosen as it imposes a limit on the daily

access and a charge for the extra usage (“Pricing for maps, routes, and places”, 2019).

Finally, the system development is open to utilizing third-party software libraries.

However, pre-testing those libraries should be made to ensure that they can work properly as

expected and meet the system development needs.

5.2.2 Software for system prototype development

Client-side software

The client is developed based on three web technologies. Specifically, Hyper Text

Markup Language (HTML) is used to define the arrangement of User Interface (UI) components,

Cascading Style Sheets (CSS) is used to style the appearance of UI components such as their

positions, heights, widths, colors and fonts, and JavaScript is used to define user interactions

with UI components, such as clicking on a button or submitting a form.

Many software and tools from HTML, CSS and JavaScript are utilized for this

development. Bootstrap, a gridding framework, is utilized to support quick prototyping of the

client. With a set of snippets of HTML, CSS and JavaScript, the Bootstrap framework offers

readily available utilities for arranging and styling UI components and defining interactions with

UI components. The Bootstrap framework also supports responsive design which means that the

system client built from it is able to adapt to different screen resolutions.

A JavaScript library, OpenLayers, is used to implement the interactive map and mapping

functionalities. It handles the development of map presentations, controls, and interactions.

GeoJSON is the format that OpenLayers uses to render vector map layers (i.e. field, stream and

86

boundary map layers). Multiple map layers and overlays can be grouped by OpenLayers in an

assigned order. Based on the attributes of a feature, OpenLayers can style features with different

visibilities, colors and levels of transparency. OpenLayers also provides several default map

controls including zoom in/out, pan, direction, rotation, center and full screen view, and some

default map interactions such as select, draw and hover. New controls and interactions can be

customized to specific map interaction tasks. For example, a draw interaction is customized in

this study for selecting multiple map features as a group to assign BMPs to multiple fields.

HighChart, which is a chart visualization tool, is used to present BMP evaluation results

(i.e. phosphorous, nitrogen, sediment, flow, cost, and revenue) and cost-effectiveness trade-off

curve of BMP policies. The chart is interactive and allows users to explore information by

selecting and hovering.

A visualization tool, sigma.js, is used to visualize the communication network for the

administrator to monitor the communications among users of the BMP planning subsystem. This

library can render undirected, unidirectional or bidirectional graph/network. Interaction methods

such as selecting and hovering are supported by sigma.js for users to examine nodes and their

neighbourhoods in the communication network.

Because multiple JavaScript libraries are used for the client development, the study uses a

tool, node.js, to manage JavaScript libraries. This tool helps organize the libraries in a structured

manner in the development environment. It also helps install dependences required by those

libraries. Furthermore, this tool offers version control to support updating libraries.

87

Server-side software

Golang is a programming language used for developing the server. The server handles

multiple tasks including serving static files such as images and HTML templates, routing and

redirecting HTTP requests from the client, processing information, and reading and writing

databases. In recent years the Golang community has developed several libraries and packages to

support the server development and enrich the functionalities of Golang. In this study, a Golang

package, Gomail, is used to enable email communications among stakeholders in both

administration and BMP planning subsystems.

To facilitate user tasks and fulfill their information requirements, three external models or

tools are called from the server. Firstly, an integrated economic-hydrologic model is called from

the server to generate the modelling results for the “What if” BMP scenarios. Secondly, an

optimization model is called from the server to generate the BMP policy/management

information. Finally, a tool “Pandoc” is called from the server to generate PDF reports. Pandoc is

an executable tool for converting a document into other formats.

For data management, the system development uses two databases. The database

PostGRE is used to manage administration-related and scenario-related data, such as user

account and user scenario information. The database SQLite is used to store BMP planning

information such as the economic cost and environmental benefits generated from BMP scenario

evaluation and BMP policy/management information generated from scenario optimization.

Table 5-1 provides a summary of the software for the system development, their

functions, and corresponding system/subsystem and client/server.

88

Table 5-1: Software for system development

Software Functions System/subsystem Client/Server

JavaScript

HTML

Bootstrap

CSS

Node.js

Highcharts

OpenLayers

Sigma.js

User interaction/information processing

UI components arrangement

Gridding framework

UI component style

Library management

Chart visualization

Online mapping library

Network visualization

All

All

All

All

All

Planning subsystem

Public/Planning subsystems

Administration subsystem

Client

Golang

PostGRE

SQLite

Integrated model

Optimization tool

Pandoc

HttpRequest handling, Information

processing

Data management (Administration,

Scenario)

Data management (BMP planning)

BMP evaluation

BMP optimization

PDF generator (external)

All

All

Planning subsystem

Planning subsystem

Planning subsystem

Planning subsystem

Server

5.2.3 System localization for the Gully Creek watershed

The system localization refers to a process wherein the system is customized to a specific

location, which is, in this case, the Gully Creek watershed. Table 5-2 lists the tasks by

subsystems and then by modules required for the system localization for the Gully Creek

watershed.

In the “Information sharing site” module of the public subsystem, the Gully Creek

watershed GIS layer is set up as the interactive map to provide the geospatial context for public

annotating. In “Public information center” module, the public information on agri-environmental

policies and BMP related technical knowledge is collected and uploaded into the database. For

89

the Gully Creek watershed, the collected public information is mostly associated with Ontario

and the Great Lakes region.

Table 5-2: System localization checklist

Subsystems Module Description

The public subsystem Information sharing site Render Gully Creek watershed map to set up geospatial context for annotations

Public information center Maintain agri-environmental policies and BMP related technical knowledge in the public information table

The BMP planning subsystem Scenario exploration 1. Provide the Gully Creek watershed map

2. Prepare the integrated economic-hydrologic model for BMP evaluation in the Gully Creek watershed

Policy/Management 1. Provide the Gully Creek watershed map

2. Prepare the optimization model for spatial targeting of BMPs in the Gully Creek watershed

Report A basic description of the Gully Creek watershed and headings for BMP evaluation results in the “What if” BMP scenarios and BMP spatial targeting in the “Policy/Management” scenarios for the HTML template (for HTML) and the report text template (for PDF)

The administration subsystem User registration Prepare the farm number and field number lookup table to match farmer users with their fields

In the “Scenario exploration” module of the BMP planning subsystem, the Gully Creek

watershed GIS layers is set up as an interactive map for users to develop BMP scenarios and

view the modelling results. The engine of the “Scenario exploration” module is the integrated

hydrologic-economic model for the Gully Creek watershed, which was developed by the Water

Evaluation Group (WEG) at the University of Guelph (Shao et al., 2017). The basic datasets for

the integrated modelling include geospatial data (DEM, land use, soil), climate data, flow and

water quality data, BMPs and agricultural management data, and related agricultural economic

90

data. Within the integrated modelling, the farm economic modelling was calibrated and validated

based on checking magnitudes of model simulation against various sources of literature

including government reports such as Ontario Ministry of Agricultural, Food and Rural Affaires

(OMAFRA) crop budgets. The SWAT hydrologic model was calibrated through adjusting model

parameters to optimize the agreement between water quantity and quality monitoring data and

model simulation results (Shao et al., 2017). The integrated modelling generates information on

economic costs, environmental benefits, and cost-effectiveness of various “What if” BMP

scenarios.

In the “Policy/Management” module of the BMP planning subsystem, an optimization

model is prepared to generate BMP policy/management information based on field-specific BMP

cost and environmental benefit data for the Gully Creek watershed (Shao et al., 2017). The BMP

policy/management information includes the spatial configuration of selected fields and BMP

types, and also economic costs and environmental benefits resulted from the configuration. The

information is displayed in the interactive map and the chart.

In the “Report” module of the BMP planning subsystem, a HTML template and a text file

template for the Gully Creek watershed are prepared for generating HTML and PDF reports

respectively. Both templates include a background chapter about the watershed such as climate,

land use, slope pattern, soil, and BMPs. BMP evaluation results from “What if” BMP scenarios

and “Policy/Management” scenarios are inserted into the templates to generate reports.

For the administration subsystem, a look-up table on farm number and field number is

used for the “User registration” module. A dropdown list is set up based on the lookup table for

the administrator to select the farm number to register a user.

91

5.2.4 A demonstration of the system prototype functionalities

This section provides a demonstration of the system prototype functionalities. The

demonstration illustrates how the information model is fulfilled by system functionalities and

user interactions. The descriptions are provided for each subsystem and then modules within the

subsystem.

5.2.4.1 The public subsystem

The public subsystem is developed to support the communication process of the public

information. A welcome webpage is developed to facilitate users to access the two modules of

the public subsystem (Figure 5-2). The left panel is used to access the information sharing site,

and the right panel is used to access the public information center.

Figure 5-2: The welcome webpage of the public subsystem

Information sharing site Pubic information center

92

Module: Information sharing site

The layout of the information sharing site is shown in Figure 5-3. The map selector above

the interactive map is for users to select the map layer of the Gully Creek watershed. The Gully

Creek watershed is rendered on the interactive map to provide a geospatial context for submitting

and viewing public annotations. The annotation list is created beside the map to display all the

submitted annotations in the Gully Creek watershed. The annotation filter tool is placed above

the list to allow users to filter annotations based on annotation contents or categories.

Figure 5-3: The interface of the information sharing site

To support users to submit public annotations, a web form is implemented for users to

submit information along with annotations (Figure 5-4). The information content includes soil

condition, crop type, environmental concerns, land management practices, structural

management practices, and others. The form also allows users to attach a picture to the

annotation. The form can be extended to include other topics if necessary.

93

Figure 5-4: The form for uploading information on field characteristics and BMP adoption

On the interactive map of the Gully Creek watershed, the submitted annotations are

presented on both the map and the annotation list. The annotations in the list are linked to the

annotation markers on the map through their shared geo-locations. Users can either interact with

the map or use the annotation filter to explore the submitted annotations:

1) The interactive map allows users to select an annotation by hovering over an

annotation marker. When an annotation marker is selected, the marker and its corresponding

annotation on the annotation list will be highlighted to show the information content.

2) The annotation filter above the annotation list allows users to select annotations based

on information contents or categories. For example, users can filter annotations by management

Web form

Click

94

practices. When the filter is applied, all the annotations including the specified practice will be

highlighted on both the map and the list.

Module: Public information center

The public information center provides a searching tool for farmers to use a keyword to

search for agri-environmental policies and BMP related technical knowledge. The searching

results will be presented in a result table that comprises of three columns (Figure 5-5). The left

column displays information on agri-environmental policies, the middle column presents web

contents related to BMPs, and the right column shows BMP technical documents.

Figure 5-5: The interface of the public information center

95

5.2.4.2 The BMP planning subsystem

The BMP planning subsystem is developed based on the information model to support

the communication process of the BMP planning information. The subsystem supports farmers

and conservation managers to evaluate the economic and environmental effects of BMPs

including conservation tillage, cover crop, nutrient management and WASCoBs. The subsystem

also supports conservation managers to design BMP policies subject to environmental targets or

economic constraints.

Module: Access control

To support user access control, a system login page is created for users including farmers

and conservation managers to enter their user name and password for authorization (Figure 5-6).

When the user authentication succeeds, the user will be directed into the BMP planning

subsystem and start to use the “Scenario exploration” module.

Figure 5-6: The login webpage of the BMP planning subsystem

User name

Password

96

Module: “Scenario exploration”

The scenario exploration supports farmers and conservation managers to evaluate

economic and environmental effects of agricultural BMPs. The task starts with creating a “What

if” BMP scenario. A form is implemented for the user to input a scenario name and a scenario

description (Figure 5-7). When the information on scenario name and description is submitted, a

“What if” BMP scenario is created, and an interface will be displayed for scenario development.

Figure 5-7: The form for creating a “What if” BMP scenario

Figure 5-8 shows the WebGIS interface for scenario development. On the interactive

map, the farm field and WASCoB layers are rendered. The field map layer is used to assign three

land management BMPs - conservation tillage, cover crop and nutrient management to fields,

and the WASCoB layer is used to assign the structural BMP, WASCoBs, to specific locations.

To support the BMP assignments, two button groups are provided above the interactive map. The

right button group allows users to select a BMP type, and the left button group provides two

feature selection methods (i.e. individual or multiple selection) for users to assign a BMP to an

individual or multiple fields or locations. The BMP assignment table on the right of the map is

developed to show the current BMP assignments in fields and will be updated when a new

assignment of BMPs is made. The CC, CT and NM in the table represent Cover Crop,

Scenario name

Scenario description

97

Conservation Tillage and Nutrient Management respectively. Below the BMP assignment table

are two buttons. The “Reset” button is created for resetting or clearing the current BMP

assignments. The “Proceed” button is used to save the BMP assignment table to the PostGRE

database and send the BMP assignment information to the integrated economic-hydrologic

modelling for evaluating the developed “What if” BMP scenario.

Figure 5-8: The WebGIS interface for developing a "What if" BMP scenario

The WebGIS interface allows a user to review the assignments of BMPs by selecting the

four BMP types. As shown in Figure 5-9, when a BMP type is selected, the map will be updated

to show BMP assignment fields or locations for the specified BMP type. For conservation tillage,

cover crop and nutrient management, the fields with BMP assignments are highlighted in orange.

For WASCoBs, the locations with WASCoB assignments are highlighted with a light-yellow

border for the dots.

98

Figure 5-9: BMP assignments in the Gully Creek watershed

After scenario development, the integrated economic-hydrologic model will be triggered

to generate economic and hydrologic modelling results in the watershed. The hydrologic

modelling results are generated by the SWAT model, which include phosphorus, nitrogen, and

sediment loadings and flow from each field in the watershed. The economic results are generated

by a farm economic model for three economic variables, i.e. net return, revenue, and production

cost at the field level. When the scenario is evaluated, the economic and hydrologic modelling

results will be written into a SQLite database and at the same time, a GeoJSON file will be

generated for presenting the modelling results on the interactive map. In this system prototype,

the BMP scenario is evaluated for a ten-year time frame from 2002 to 2011 to understand the

long-term effects of BMPs on economic and environmental variables.

Conservation tillage Cover crop

Nutrient management WASCoB

99

Figure 5-10 shows the WebGIS interface for presenting the modelling results for the

“What if” BMP scenario, which involves three economic variables (i.e. net return, revenue and

cost) and four environmental variables (i.e. phosphorous, nitrogen, and sediment loadings and

flow). The presented results include both on-site and off-site results. The on-site results refer to

the results at a field scale; the off-site results refer to the accumulative results at the watershed

outlet. To support information exploration, a button group above the map allows users to check

out the results for a specific environmental or economic variable. For example, when the button

“Phosphorus” is clicked as shown in the Figure 5-10, the map is updated with the “total

phosphorus” results. Because the evaluation period is ten years, the value of total phosphorus on

each field is an averaged value for the ten years. The total phosphorus at the off-site level

represents the result at the outlet of the watershed.

Figure 5-10: The WebGIS interface for scenario evaluation result presentation and exploration

100

The evaluation chart is developed to present the scenario results of a specific variable

over a time period at on-site and off-site levels. When a field on the interactive map is selected,

the evaluation chart shows the on-site results of the selected variable (e.g. total phosphorus) for

that field. When all the fields on the map are deselected, the chart will render the off-site results

of the selected variable (e.g. total phosphorus) at the watershed outlet. For example, the

evaluation chart in Figure 5-10 shows the off-site results of total phosphorus over a ten-year time

frame from 2002 to 2011. The curved line in the chart indicates the results of total phosphorus in

each year, and the straight line shows the averaged result of total phosphorus over the years.

After scenario evaluation, the module supports comparing the BMP scenario with a

baseline scenario. Figure 5-11 shows the WebGIS interface for scenario comparison. A scenario

list is provided for user to select a baseline scenario for the comparison, which can be either

historical or conventional scenario. When a baseline scenario from the scenario list is chosen, the

comparison results will be rendered on the interactive map.

101

Figure 5-11: The WebGIS interface for scenario comparison

Above the interactive map, a button group is provided for users to examine the scenario

comparison results for specific variables and the cost-effectiveness of BMPs. The comparison

results include three groups: economic costs, environmental benefits, and cost effectiveness of

BMPs. The economic costs and environmental benefits are estimated through comparing the

evaluation results of the two scenarios (i.e. the “What if” and a baseline scenario). The economic

results include the differences of the three economic variables, i.e. cost, revenue and net return,

and reflect the economic effects of BMPs in the fields. The environmental results include the

differences of the four environmental variables i.e. phosphorus, nitrogen, and sediment loadings

and flow, and reflect the environmental effects of BMPs. Figure 5-12 shows the comparison

102

results of the net return in the fields between the “What if” and the baseline scenario. The

negative net return implies the BMP costs incurred by implementing the BMPs. The positive net

return implies economic gains from implementing BMPs (e.g. cost saving from fertilization

management). Similarly, Figure 5-13 shows the comparison results of total phosphorus in the

fields between the “What if” and the baseline scenarios. The negative number indicates the

phosphorus reduction from implementing the BMPs.

Figure 5-12: Differences in net return between the “What if” and the baseline scenario

103

Figure 5-13: Differences in total phosphorus between the "What if" and the baseline scenario

The integrated result is the combination of both economic and environmental results,

which represents the cost-effectiveness of BMPs. The cost effectiveness for phosphorus,

nitrogen, and sediment loadings and flow are estimated by the differences of the specific

environmental variable divided by the differences of the net returns between the “What if” and

the baseline scenario. The BMP cost effectiveness is represented as the environmental changes

for $1,000 BMP cost. The units for BMP cost-effectiveness are mm/$1,000 for flow, ton/$1,000

for sediment yield, and kg/$1,000 for TN and TP yields, which indicate the water

quantity/quality effects per $1,000 BMP costs (Yang et al. 2013).

Figure 5-14 shows the cost-effectiveness of BMPs on the total phosphorus reduction. The

pattern of cost-effectiveness should be interpreted with caution because the cost effectiveness

results may have negative or positive signs due to various combinations of water quantity/quality

104

effects and BMP costs (Table 5-3). In typical cases, environmental benefits (baseline – “What

if”) are achieved with economic loss (negative net return), which lead to negative cost-

effectiveness (Situation 1). The higher absolute value of cost-effectives indicates more cost-

effectiveness. In some cases, environmental benefits could be achieved with economic gains (net

return increase), for example, when nutrient management reduces the cost on fertilizer (Situation

2). However, due to complex nutrient cycling and hydrologic processes and landscape

conditions, BMP implementation could also lead to environmental harms (Situation 3 and

Situation 4). Those cases need to be specifically analyzed (Yang et al. 2013).

Figure 5-14: Cost-effectiveness of BMPs on total phosphorus reduction

105

Table 5-3: Various combinations of water quantity/quality effects and BMP costs (Yang et al. 2013)

Situation Environmental benefits Economic costs Cost-effectiveness

(integrated)

1 Water/pollutant reduction (+) Net return reduction (-) Negative (-)

2 Water/pollutant reduction (+) Net return increase (+) Positive (+)

3 Water/pollutant increase (-) Net return reduction (-) Positive (+)

4 Water/pollutant increase (-) Net return increase (+) Negative (-)

In the WebGIS interface for scenario comparison, there is a button group below the

scenario list. The “Optimization” button allows users to enter the “Policy/Management” module,

and the “Report” button allows users to generate reports based on the scenario comparison

results.

Module: Policy/Management

The “Policy/Management” module supports users to design two types of policies: the

environmental policy/management, which minimizes economic costs subject to environmental

targets and the economic policy/management, which maximizes environmental benefits subject

to financial constraints.

The WebGIS interface is developed to allow users to set parameters for running the

optimization model for the BMP policy/management scenario (Figure 5-15). The required

parameters include the policy/management type, the fields for optimization, the pollution type,

and the BMP policy/management constraint. For an environmental policy, this constraint is a

range of environmental benefit targets (e.g. phosphorus reduction in kg/year). For an economic

policy/management, this constraint is a range of the BMP costs in dollar/year. The

policy/management type, the farm fields for optimization, and the pollution type are prerequisites

to determine the BMP policy/management constraint.

106

To support users to set the parameters, the WebGIS interface provides button groups,

buttons and an input box. Above the interactive map, there are three button groups and a button.

The left button group allows users to select a policy/management type – “Env” for the

environmental policy/management and “Eco” for the economic policy/management, the middle

button group allows two field selection methods for users to specify the farm fields in which the

BMP policy/management will be applied (“S” for a single field and “M” for multiple fields), and

the right button group allows users to select a pollutant type (P – Phosphorus, N – Nitrogen, S –

Sediment, F – Flow), indicating which pollutant the policy/management is applicable to. The red

button “RS” beside the right button group is used to reset the policy/management type and field

selection.

Figure 5-15: The WebGIS interface for the “Policy/Management” module

After specifying the policy/management type, selected fields (multiple fields/locations or

all crop fields/locations) and the pollution type (such as phosphorus), the user can obtain a

107

default range for BMP policy/management constraint (corresponding to no BMP and full

implementation of BMPs in the watershed) by clicking the yellow "Range" button which is

above the optimization table. This default range indicates the upper and lower limits of the

environmental benefits under the environmental mode or the upper and lower limits of the BMP

economic costs under the economic mode. For example, when the environmental policy type and

the total phosphorus (TP) pollution type are selected and all the fields are selected for

optimization (Figure 5-16), the default environmental benefit range for BMP policy/management

constraint will be 0 – 2,747.54 kg/year which indicates the minimum and maximum TP

reductions in the selected fields. Similarly, when the economic policy/management type and the

TP pollution type are selected and all the fields are selected for optimization, the default BMP

economic cost range for policy/management constraint can be -20013.53 – 45857.27 $/year,

indicating the minimum and maximum BMP costs on the selected fields. The negative BMP cost

means net return gains (positive net return change), which may be caused by nutrient

management BMP in terms of fertilizer reduction (reduced production costs) under equivalent

yield and revenue. The positive BMP cost means net return losses after BMP(s) application.

Figure 5-16: The default range of BMP policy/management constraints

0 - 2747.54

-20013.53 - 45857.27

Environmental policy type + total phosphorus type

Economic policy type + total phosphorus type

Range of TP reduction (kg/year)

Range of BMP cost ($/year)

108

The default BMP policy/management constraint is displayed in the input box above the

optimization table and on the left of the “Range” button. Based on the default BMP

policy/management constraint, the user can revise the default range to specify a new BMP

policy/management constraint, such as the range of environmental targets (pollutant reductions)

or the range of financial constraints (BMP costs). After the setup of the BMP policy/management

constraint, the user can click the “Opt” button next to the “Range” button to run the optimization

model to obtain BMP policy/management optimization results.

Figure 5-17 and Figure 5-18 show the interfaces for presenting and exploring the

environmental and economic policy/management information. The environmental

policy/management information is generated using the parameters in Table 5-4.

Table 5-4: The parameters for producing the BMP policy/management information

Policy/management

type

Selected fields Pollutant type BMP policy/management constraint

Environmental See Figure 17 Phosphorous 2,200.00 – 2,747.54 kg/year phosphorus

reduction (indicating a 30% - 37%

phosphorus reduction - 37% is the

percentage when the maximum reduction

2,747.54 kg/year phosphorus is obtained)

Economic See Figure 18 Phosphorous -20,013.53 – 40,000.00/year BMP cost

The layouts of two interfaces are identical. The BMP policy/management information is

visualized using three information presentation options. The interactive map is implemented to

show BMP combinations in each farm field. Different color indicates different combinations of

BMPs. The information on field-specific BMP combinations is also presented in the

policy/management table. Each row of the table indicates how BMPs are assigned to a specific

field (“Y” means assigned and “N” means not assigned). Moreover, a line chart is used to present

109

the trade-off relationship between the economic costs and environmental benefits within the

range of constraints. In both Figures, i.e. Figure 5-17 and Figure 5-18, two zoomed-in

screenshots are provided to facilitate the examination of the chart. The Y-axis of the chart

indicates the environmental benefits and X-axis indicates the economic costs. Ten points on the

curve represent ten sets of BMP configurations on the selected fields with the corresponding

economic costs and environmental benefits. When clicking a point on the chart, the

corresponding BMP configuration will be rendered on the map and the policy/management table.

Figure 5-17: The interface for exploring environmental policy/management information

Optimization table

Interactive map

Optimization chart

WASCoB

110

Figure 5-18: The interface for exploring economic policy/management information

Module: Discussions between farmers and conservation managers

The interface for supporting communications between farmers and conservation

managers is shown in Figure 5-19. The discussion topic form is provided to allow farmers to

submit various discussion topics. When a topic is created, the topic is appended to the discussion

topic list. The topics on the list are clickable to expand the discussion thread of the topic in a

discussion window. The discussion window enables users to provide replies or comments to the

topic. Conservation managers can access and reply to all the discussion topics created by

farmers. An email function is integrated with the discussion function. When farmers and

conservation managers submit discussion topics or post on the discussion thread, emails will be

sent to them for notification.

Optimization table

Interactive map

Optimization chart

WASCoB

111

Figure 5-19: The interface for supporting communications between farmers and conservation managers

Module: Scenario report

To facilitate communicating the “Scenario exploration” scenario and

“Policy/Management” scenario results, both HTML and PDF reports can be generated from the

system (Figure 5-20). The HTML report can be generated after running the scenario and it allows

for a quick review of the scenario results. Because the HTML report is essentially a webpage, all

the components on the HTML report are interactive. For example, users can click on buttons on

the HTML report to check the BMP evaluation results for different variables on the map.

The PDF report shares the same report content with the HTML report. It can be

downloaded, printed, and saved as a file. A template is used to align information contents in the

PDF report.

Discussion topic

Discussion topic list

Discussion thread

112

Figure 5-20: Scenario reports in HTML(Left) and PDF(Right) formats

To enable the report generation, the system provides two separated buttons in both

“Scenario exploration” and “Policy/Management” modules. One is for HTML report generation,

and the other is for PDF report generation. When the button is clicked, the report will be

generated, and the new report will be displayed in a new browser tab.

5.2.4.3 The administration subsystem

The administration subsystem shares the same login page with the BMP planning

subsystem (Figure 5-6). It requires system administrator’s user name and password for the login.

When the user authorization completes, a webpage will be displayed for the administrator to

view the monitoring information (Figure 5-21). On the up-right corner of the page, three tabs are

provided for the system administrator to navigate different system modules and functions: the

HTML PDF

113

“System” tab to view the monitoring information in tables, the “Network” tab to view the

monitoring information through the communication network, and the “Registration” tab to

register new users to use the BMP planning subsystem.

Figure 5-21: The webpage after login to the administration subsystem

Module: System usage monitoring

The administration subsystem provides two tables and a communication network to

support the system administrator to monitor the use of the BMP planning subsystem by farmers

and conservation managers. The tables include a usage summary information table and a

personal usage information table (Figure 5-22). The usage summary information table shows the

summarized usage information of the system, including the total number of users, scenarios, and

discussion requests that users have created, while the personal usage information table displays

114

usage information by individuals, including user names, the number of scenarios and discussions

created by each user, and scenario completion status that is indicated by the rate of completion.

Figure 5-22: Tables for displaying the usage information of the BMP planning subsystem

The communication network visualizes communication relationships among farmers and

conservation managers based on their discussions in the BMP planning subsystem (Figure 5-23).

On the communication network, each node represents a user (farmer/conservation manager), and

the diameter of a node reflects the centrality closeness of that user, indicating the frequency the

user communicates with others. A connection on the network represents an existing connection

between two users. Links have two colors given that each user has one of the two roles, e.g

farmer or conservation manager. Links between farmers are colored in yellow and those between

conservation managers and farmers are colored in red. A user list is also provided on the left-

hand side of the communication network to show all the farmers and conservation managers in

the network.

115

Figure 5-23: Communication network for displaying communication information among farmers and conservation managers

Interaction with the communication network can go through either the user list or the

communication network. Clicking on a specific user in the list or in the communication network

will highlight the user and his/her neighbours as well as their connections on the network. Upon

selecting a user, the administrator can also send messages to the user through an email messaging

tool (Figure 5-24).

116

Figure 5-24: Interaction within the communication network

Module: User registration

A registration form is provided for the system administrator to register new users. The

required information for registration includes username, email address, role and farm number

(Figure 5-25). The role of the user can be either a farmer or a conservation manager. The farm

number is used to link a farm to its fields to control the user access to the fields within the farm.

The farm number for the conservation managers is 0 by default so they have access to all fields

in a watershed. Because the user registration involves preparing a copy of all data and models in

the BMP planning subsystem for the created user, the registration takes a few minutes. Once the

registration is completed, an email will be sent to the user to notify the compeletion of

registration. A self-assigned username and an encrypted password is provided through the email

and only known to the registered user.

117

Figure 5-25: The user registration form

5.3 Summary

This chapter introduces a prototype of the WebGIS-based decision support system for the

Gully Creek watershed. The chapter starts with introducing the study area. The “System

localization” section discusses how the WebGIS-based decision support system is adapted to the

Gully Creek watershed. The localization procedure shows the implementation of the system

design to a specific watershed. After that, each subsystem and its modules are demonstrated from

three aspects: user interface, user interactions and information presentations. The demonstration

of the user interface focuses on the layout of the interface, the demonstration of user interactions

shows the actions for facilitating user tasks within the system, and the demonstration of

information presentations emphasizes the methods for information presentation and interactions

with the information content. The system demonstration illustrates the fulfillment of information

needs of stakeholders and the processes of information communications based on GIS, models

and ICTs.

118

Chapter 6 Evaluating the WebGIS-Based Decision Support System

This chapter presents the evaluation of the WebGIS-based decision support system

prototype for the Gully Creek watershed. The content of this chapter is organized into three

sections. The first section defines usability evaluation. The second section introduces evaluation

methods, including both evaluation by direct use and evaluation during demonstration. In the

third section, evaluation results are presented.

6.1 Usability evaluation

6.1.1 Defining system usability

The term “usability” has been widely used in the studies of human-computer interaction

(HCI) and graphical interactive interface to describe and measure the performance of a system

and how “usable” it is likely to be in a specified context of use (Fischer, 2001). However,

challenges exist in defining the term “usability” for its complexity and specific context. As

indicated by Issa & Isaias (2015), “usability is not determined by just one or two constituents but

is influenced by a number of factors which interact with one another in sometimes complex

ways”. The usability, when being investigated, is factually a collective outcome of several

variables of a system including system function, characteristics of users and tasks, and their

inter-relationships (Fischer, 2001).

The two widely accepted definitions of usability are from ISO 9241-11: “the extent to

which a product can be used by specified users to achieve specific goals with effectiveness,

efficiency and satisfaction in a specified context of use” and ISO 9126-1 : “the capability of the

software product to be understood, learned, operated, attractive to the user, and compliant to

119

standards/guidelines, when used under specific conditions”. Based on the HCI and software

engineering perspectives, both definitions outline the key elements of the usability as

effectiveness, learnability, efficiency and satisfaction. The effectiveness measures if the system

could support users to successfully achieve their objectives in compliance with what the system

is designed for, the learnability indicates to what extent the system is easy for user to become

familiar with and know how to use it on subsequent visits, and the efficiency and satisfaction

denote the time spent to complete certain tasks and whether users enjoy interacting with the

system during the process.

6.1.2 Usability evaluation approaches

Evaluating the usability of a system has become a critical step for releasing the system

for practical use. The approaches for usability evaluation can be classified into two categories,

namely the quantitative approach and the qualitative approach (Grinnell, 2001). A major

difference between the two approaches lies in the data that they use for the evaluation.

Specifically, the quantitative approach relies on the collection and analysis of numerical data,

whereas the qualitative approach relies on the interpretation of descriptive text/narrative

(Grinnell, 2001). The data for evaluation can be collected from different sources. The

quantitative numerical data is commonly collected from survey or questionnaire; the qualitative

descriptive data can be collected through participant observation, open-ended questions,

interview, or focus group (Garbarino & Holland, 2009). The focus group is a form of group

conversations where key users are invited to discuss their opinions towards a central topic

together (Mazza & A. Berre, 2007).

120

Both quantitative method and qualitative approaches have specific advantages. The

quantitative approach, due to its numeric basis, can apply statistical techniques to analyze

numerical data and reveal the correlations between evaluation measures and evaluation

objectives. Hence, the quantitative approach has been widely used to address questions like to

“what” extent a factor can lead to a certain result (Barkmann et al., 2009). Compared to the

quantitative approach, the qualitative approach seeks to explain the “why” and “how” behind the

“what” (Kaplan & Maxwell, 2006). Through interpreting the text/narrative, the qualitative

approach provides in-depth descriptive information about evaluators’ feelings, experiences and

perspectives regarding the system use (Fossey et al., 2002). Instead of building direct links

between evaluation measures with evaluation objectives, the qualitative approach explains

behaviours and outcomes in the complex social-technological context.

In this study, the qualitative approach was selected for evaluating the WebGIS-based

decision support system. The selection was based on two considerations. Firstly, the evaluation

mainly focuses on the process of system use by evaluators. The qualitative approach helps to

probe information to explain how the WebGIS-based decision support system fulfills users’ task

and information requirements. Secondly, the implemented system is a prototype. The qualitative

approach can provide a great opportunity to collect suggestions on further improving the system.

As indicated by Kaplan & Maxwell (2006), using qualitative method can help in identifying

potential problems, thereby providing opportunities to improve the system as it develops.

However, it is worth noting that the qualitative approach for evaluation has its limitations. The

qualitative approach is meant to study a specific issue (i.e. BMP adoption) in a particular context

(i.e. the Gully Creek watershed). The results may be not generalizable to other areas with

different geographical conditions and social, cultural and economic contexts (Leung, 2015).

121

6.2 Evaluating the WebGIS-based decision support system using a qualitative approach

Evaluating the WebGIS-based decision support system involved two steps. Firstly,

usability evaluation measures were identified. Secondly, usability evaluation methods were

introduced to illustrate how the evaluation was conducted and data were collected.

6.2.1 Usability evaluation measures

Nielsen’s heuristics presents a classical list of usability principles (Nielsen, 1990). It

classified common design issues into ten categories, including visibility of system status, match

between system and the real world, user control and freedom, consistency and standards, error

prevention, recognition rather than recall, flexibility and efficiency of use, aesthetic and

minimalist design, help users recognize, diagnose and recover from errors, help and

documentation. By far, Nielsen’s heuristics have been referenced by many studies (Mankoff, et

al., 2003; Pinelle, Wong & Stach, 2008). However, due to different characteristics of systems,

the list has to be extended or modified to meet specific usability evaluation objectives. For

example, Mankoff et al. (2003) developed a modified set of principles for evaluating ambient

displays, which are abstract and aesthetic peripheral displays portraying non-critical information

on the periphery of a user’s attention. They referred to Neilsen’s heuristics but eliminated the

non-applicable ones due to the passive nature of the displays. Moreover, Pinelle, Wong & Stach

(2008) developed a set of principles for evaluating the video game design. They stated that while

the Nielsen’s heuristics cover generic design issues, they are not specific enough to address some

important issues in game video design, such as “using proper camera angles when displaying the

game world or providing intuitive control mappings”.

122

In this study, the usability evaluation measures were developed based on the updated

Information System Success Model (Figure 6-1) by DeLone and Mclean (2003), which was

originated from Nielsen’s heuristics (Nielsen, 1990). The Information System Success Model

provided a framework to understand different aspects of a “successful” information system,

including not only the usability principles that focus on system interface but also factors related

to information and services. The model also offered a mechanism for organizing the evaluation

measures. According to the model, evaluation measures were organized into six categories, i.e.

system quality, information quality, service quality, satisfaction, intention to use, and impact

(Table 6-1).

The system quality characterizes the technical success of the system. Evaluation

measures in this category include user interface, task completeness, and learnability. User

interface determines how users communicate with system information, functions, and modules.

A good user interface should be clear, simple, direct, easy to navigate, and enable users to fully

control their actions (Nielsen, 1990). Task completeness means that the system can successfully

complete required user tasks to fulfill their objectives (Hamilton & Chervany, 1981). Lastly,

learnability indicates the easiness with which the system can be picked up and understood by

users.

123

Figure 6-1: The updated Information System Success Model (DeLone & Mclean, 2003)

The information quality describes the success of information in supporting BMP

adoption. It can be measured by comprehensiveness, accuracy, clarity, and interactivity. The

comprehensiveness means that the information provided to users should be complete and

adequate; the accuracy means that the information should be correct; the clarity means that the

information should be clear in meaning; and the interactivity means that, for complex

information content, the presentation methods should enable users to explore information in

various dimensions and in multiple ways.

The service quality characterizes the success of web services. This category includes two

evaluation measures: service responsiveness and support. The service responsiveness describes

the timeliness of web services responding to the service requests. For web-based systems, the

service responsiveness is significantly related to the user experience with the system (Palmer,

2002). The support refers to the availability of supporting materials for users to complete their

tasks.

124

Table 6-1: Measures for evaluating the WebGIS-based decision support system

Category Measures Description

System quality User interface The system interface is simple, clear, easy to

navigate, and enables users to fully control

their actions.

Task completeness The system can successfully complete

required user tasks for BMP adoption.

Learnability The system is easy to be picked up and

understood by users.

Information quality Comprehensiveness Information provided to users is complete

and adequate.

Accuracy Information is correct.

Clarity Information is clear in meaning.

Interactivity Information is interactive for users to explore

different dimensions of information.

Service quality Responsiveness The system responds to user’ service requests

in a timely manner.

Support The system provides supporting materials to

facilitate and guide users to complete tasks.

User satisfaction Satisfaction Users are overall satisfied with the system.

Intention to use Intention to use Users are willing to adopt the system in their

future work.

Impact Awareness The system increases user’ environmental

awareness.

Knowledge The system improves user’ knowledge of

BMPs.

The satisfaction describes the degree of user satisfaction based on their perceived system,

information, and service quality.

The intention to use describes the willingness of users to adopt the system in the future.

The intention to use and user satisfaction are closely interrelated. When users have satisfied

experience with using the system, their intention to use the system will be strengthened. They are

125

more likely to re-use the system and recommend the system to others. On the other hand,

dissatisfaction weakens their intention to use the system and reduces the usage of the system.

The impact measures perceived usefulness and challenges of the system for supporting

the adoption of BMPs. The positive impact will reinforce subsequent intention to use and user

satisfaction, while the negative impact would cause a decreased use of the system and even a

discontinuance of the system use in the worst case. Evaluation measures in this category include

awareness and knowledge. Awareness means that the system can improve user’s environmental

awareness, and knowledge means that the system is able to increase user’s knowledge of BMPs

and support their decisions on BMP adoption.

6.2.2 Usability evaluation methods

In this study, usability evaluation uses two complementary methods: evaluation by direct

use and evaluation during demonstration (Figure 6-2). Evaluation by direct use is implemented

by inviting a panel of usability experts to evaluate the system. The main focus of this method is

to identify system design flaws subject to a set of usability principles such as error prevention

and simplicity. According to Jeffries, & Desurvire (1992), evaluation by direct use can work

very well to generate useful results from only 3-4 evaluators (Jeffries, et al., 1991; Togazzini,

1992). However, evaluation by direct use has its limitations. Firstly, evaluation by direct use

involves usability experts who may lack practical domain knowledge (Chin, Diehl, & Norman,

1988; Nielsen,1993; Holzinger, 2005). The usability experts mostly focus on interface

interactivity but may not emphasize on examining realistic user scenarios. For example, a

usability expert may not select those farm fields with higher erosion potential for conservation

tillage BMP. Secondly, evaluation by direct use has no built-in mechanism to ensure functions in

126

the system are thoroughly explored. As Tyllinen, et al. (2016) mentioned, evaluation by direct

use may be not able to identify missing functionalities.

To address the aforementioned challenges, this study also used evaluation during

demonstration. Complementary to evaluation by direct use, evaluation during demonstration has

several advantages. In addition to identifying the violation of usability principles, evaluation

during demonstration allows for discovering missing functionalities of the user scenario

(Tyllinen, et al., 2016). This is realized by broadening the scope of evaluator engagement.

Evaluation during demonstration allows novice users or usability non-experts to be engaged

within the evaluate process, especially the ones with practical domain knowledge. As noted by

Tyllinen, et al. (2016), demonstrations could focus evaluators on the tasks and information

without distracting or burdening them. Demonstrations allow usability non-experts to evaluate a

wide range of use scenarios and functionalities. The results from both evaluation by direct use

and evaluation during demonstration are collected and coded into the evaluation measures to

provide an overall assessment of system usability.

127

Figure 6-2: Evaluation by direct use and evaluation during demonstration

6.2.2.1 Conducting evaluation by direct use

For evaluation by direct use, four usability experts were invited: two of them were from

University IT department and the other two were research colleagues at the University of

Guelph. The two IT experts had been managing university-websites deployment for years and

had adequate experience in the domain of user interface design. The two research colleagues

were from research domains of integrated GIS and watershed modelling, and computer science.

The four experts were invited to evaluate the system using their personal computers.

These experts were asked to complete key user tasks in the subsystems as shown in Table 6-2. A

system user manual was provided to the evaluators before the evaluation and no other assistance

such as training and chauffeurs were provided. The system user manual provided a reference to

the system functions that describes the objectives and steps of each user task.

Evaluation by direct use Evaluation during demonstration

Usability principles User task and information

Usability non-expertUsability expert

Aggregated result

128

Table 6-2: Key user tasks for evaluation by direct use

Subsystem Tasks

The public subsystem Submit and load public annotations at the information sharing site

Check BMP related knowledge and local agri-environmental policies in the

public information center

The BMP planning subsystem Log into the BMP planning subsystem

Create a scenario

Develop a scenario

Evaluate a scenario

Compare the scenario with a baseline scenario

Optimize a scenario

Initiate a discussion thread by submitting a discussion topic

Submit a discussion reply on a discussion thread

Generate a HMML and PDF report

The administration subsystem Log into the administration subsystem

Check information of system usage

Explore the communication network

Register a new user account

Send a message through the communication network

After the evaluations were completed, individual interviews were conducted to collect

feedbacks from each evaluator. Based on the developed evaluation measures, their feedback was

compiled into scripts and coded into the evaluation measures (Appendix E, Table E.1). Given

that the evaluators were all with a strong technical background, the evaluation results, as

expected, were mostly related to technical aspects of the system.

6.2.2.2 Conducting evaluation during demonstration

The evaluation during demonstration was carried out in the forms of research group

discussion and program panel review. The research group discussion involved three researchers

with different areas of expertise. One researcher was from the domain of agricultural economics

129

and had extensive managerial experiences on agri-environmental programs, one researcher was

from a multi-disciplinary field of GIS and watershed analysis and developed a desktop-based

GIS system named “WhiteBox”, and one researcher was from the domain of GIS and conducted

research on a variety of GIS research topics, such as services, data modelling, visualization,

collaborative decision-making and environmental management.

The program panel review involved one system demonstration for two hours to 8 staff

members in the Ausable Bayfield Conservation Authority (ABCA) and one system

demonstration for one and a half hour to 25 people including researchers, IT staff, and agri-

environmental program managers from the Ontario Ministry of Agriculture, Food and Rural

Affaires (OMAFRA). As suggested by staff from ABCA and OMAFRA, local farmers in the

Gully Creek watershed were not included in the evaluation due to the sensitiveness in

communicating with farmers by people outside conservation authority and government. Several

staff members from ABCA and OMAFRA had experience in farming operations and understood

the BMP adoption process well. Furthermore, they indicated that the system would be preferably

used by conservation authority and/or OMAFRA staff to evaluate BMP scenarios together with

farmers when visiting them to discuss BMP adoption. Therefore, they can serve as surrogates to

evaluate the system.

Before each demonstration, a brief presentation was given to introduce the background

and objectives of the system. During the demonstration, the system was introduced in the order

of subsystems, modules and user tasks. User interactions with the system were demonstrated step

by step to illustrate how the interface design supports user tasks. The participants were asked to

evaluate the system for the interface, user task, and information (Table 6-3). To be specific, the

130

evaluation during demonstration aims to understand how participants perceive the interface

design, how they agree upon the design of task and operational process for supporting user tasks,

and how the information satisfies their needs for BMP adoption.

The research group discussion and two demonstrations were conducted using a large

projection screen and evaluators could interrupt and raise questions freely during the process.

The evaluation results from the research group discussion and two demonstrations were collected

and then, based on the developed evaluation measures, compiled into feedback scripts after the

demonstrations (See Appendix E, Table E.2).

During the evaluation by demonstration, conservation managers served as surrogates of

farmers for evaluating the system. However, as farmers are not directly involved in the

evaluation process, the evaluation method has several limitations. Firstly, conservation managers

are typically more knowledgeable and technically competent than farmers. The difficulty level of

system functions such as visualizing BMP modelling results may be underestimated. Secondly,

farmers have different social and economic characteristics. Their design of BMP scenarios may

be different from that of the conservation managers. Considering that the system aims to

facilitate farmers to make decisions on BMP adoption, it would be ideal to collect feedback from

farmers with different characteristics for evaluating the system.

131

Table 6-3: System design for evaluation during demonstration

Category Description

Interface Is the interface attractive, e.g. websites layout and color contrast?

Are the interactive operations with interface straightforward?

Are the navigations among subsystems and modules clear?

User task Is the task flow in the “What if” BMP scenario design sensible, i.e. scenario creation,

development, evaluation and comparison?

To what extent the system functions satisfy your task requirements?

Is the operation flow for conducting user tasks easy to understand and remember?

Information To what extent the information generated by the system is adequate to support your tasks

for BMP adoption?

To what extent the information is accurately and clearly presented and easy to perceive?

Are the discussion forum and Email functions useful to promote communications among

farmers and conservation managers?

6.3 The evaluation results

The results from evaluation by direct use and evaluation during demonstration were

aggregated and interpreted to develop an overall assessment of the system. Firstly, based on the

evaluation measures, the feedback scripts were examined, and those scripts associated with more

than one evaluation measures were divided into short scripts for specific measures. Secondly, all

feedback scripts were also examined to remove repetitions among the scripts. However, the

number of the repetitive scripts was recorded to indicate the significance of the opinion.

The discussions were organized by various categories of evaluation measures including

system quality, information quality, service quality, satisfaction, intention to use, and impact.

132

System quality

The evaluators provided an overall positive feedback on the system quality. They

mentioned that the system interface was user-friendly, the navigation within the system was

clear, and interactions with the system were straightforward. They also reported that the system

met the task completeness criterion, and the design of task flows was reasonable to guide

complex user tasks such as scenario evaluation and optimization.

However, there were several concerns. One major concern was regarding the learnability

of the system. The evaluators indicated that due to the complexity of the BMP scenario design

and evaluation process, farmers with less technical knowledge may have difficulties in using the

system. They suggested that “Tooltips or other assistance tools should be integrated into the

system to guide user operations and improve the learnability of the system”. Several other

concerns were reported regarding the system interface. An evaluator mentioned that “User

controls should be added when operations have dependency relationships”. An evaluator also

suggested “an improvement of the color contrast between the system background and working

space”. Furthermore, an evaluator suggested that “An adjustment of the font size should be made

to make the text more readable on the screen”.

Information quality

Evaluators agreed that the system improved information accessibility and achieved the

objective to meet various information needs by stakeholders. They, in particular, valued the

information generated from the “What if” BMP scenario evaluation and scenario optimization.

The “What if” BMP scenario evaluation provides information on the integrated economic and

133

environmental effects of BMPs; the scenario optimization identifies recommended solutions on

BMP planning (optimal BMP scenarios and their locations) subject to economic or

environmental constraints. They also mentioned that the various types of information

presentations, including the map, table, and chart, allowed them to better explore and understand

the modelling results.

The major concerns regarding information quality include the quality of data in the public

system and the information accuracy of the modelling and optimization results. Given the public

system allows farmers to submit public annotations, most of evaluators suggested that “Specific

mechanisms need to be developed to control or verify the information quality”. Because the BMP

planning information plays an essential role in support the BMP adoption decisions, evaluators

also emphasized that “BMP planning information generated from the integrated economic-

hydrologic model and optimization model needs to be accurate”, and “Explanations to the

modelling and optimization results should be added in the system to facilitate users’

understanding and interaction with the results”.

Service quality

Evaluators generally agreed that the web services were responsive, despite the fact that

the integrated economic-hydrologic model took around 2 minutes to generate the evaluation

results of BMP scenarios. Given the model simulates ten years of daily data, the running time is

acceptable to the evaluators.

One major concern regarding the service quality was about the documentation and

tutorial. Evaluators suggested that “A well-documented user manual needs to be integrated into

134

the system to explain and guide user tasks by steps, particularly for the scenario optimization.

Video tutorial would also be helpful to support the use of the system”.

Satisfaction

Based on the evaluation results on system quality, information quality, and service

quality, the evaluators were satisfied with the system overall. They valued the simplicity of the

interface design. They also assessed that the system meets the information needs for making

BMP adoption decisions. In particular, they showed a high interest in some system modules and

functions. An evaluator indicated that “The automated reporting process is very helpful in BMP

planning”. An evaluator also reported that “The embedded discussion forum and email functions

provide a convenient way for farmers to reach conservation managers”. Furthermore, an

evaluator mentioned that “The network diagram showing the communication relationships

among farmers and conservation managers is an interesting and innovative function”.

Several comments on further improvements to the design of the system were also raised.

For example, evaluators suggested that “The system can be improved by implementing an online

community forum to facilitate general discussions and sharing BMP adoption experiences” and

“The system can be improved by providing the historical BMP adoption information to

farmers”. Evaluators also suggested that “A mobile version of the public system can be

developed to improve the usability of the system”.

Intention to use

The evaluators showed a high interest in the system design and functions. Particularly,

the evaluators indicated that the WebGIS-based decision support system provides a platform for

135

communicating complex BMP modelling results to farmers and conservation managers. Staff in

the conservation authorities commented that “The system can be used when visiting farmer to

discuss BMP adoptions”. Farmers and conservation mangers can use the system to evaluation

various BMP scenarios together to develop consensus on BMP adoption. The staff in OMAFRA

also commented that “The public system has a great potential to be used for collecting data on

on-ground activities including BMP adoption”.

Impact

The evaluators indicated that the system has the potential to play a role in improving their

environmental awareness and BMP knowledge, and supporting their decisions on BMP adoption.

For example, they mentioned that “The public subsystem, specifically the public annotation

sharing site, can be used as a tool to increase farmers’ environmental awareness”. They also

mentioned that “The information center in the public subsystem is useful to find locale specific

information, such as agri-environmental policies” and “The BMP planning subsystem provides

an effective tool to improve the understanding on the economic and environmental effects of

BMPs”.

6.4 Summary

This chapter presents the evaluation of the WebGIS-based decision support system for

facilitating agricultural BMP adoption in the Gully Creek watershed. The evaluation uses two

methods including evaluation by direct use and evaluation during demonstration. The evaluation

measures include system quality, information quality, service quality, satisfaction, intention to

use, and impact. The aggregated evaluation results show that the evaluators were overall satisfied

136

with the system design and functionalities. Several suggestions on further improvements to the

system were also provided.

137

Chapter 7 Conclusion

7.1 Summary

This research developed a WebGIS-based decision support system for facilitating BMP

adoption. The system supports the information communication processes for BMP adoption.

Specifically, the system provides farmers and conservation managers with easy access to the

public and BMP evaluation and policy/management information; the system also supports

information communications between farmers and conservation managers so that they can

discuss concerns and reach consensus on BMP adoption. It is worth noting that the WebGIS-

based decision support system evaluates environmental benefits of BMPs in terms of water

quantity and quality, but not others such as soil health, ecological benefits, etc. Moreover, the

WebGIS-based decision support system does not examine the social and cultural aspects of BMP

adoption.

The Chapter 1 introduced the background and the objectives of the research. Famers and

conservation managers face various barriers on effective communications for BMP adoption

based on relevant information content, which includes field characteristics, environmental

concerns, BMP adoption status, BMP technical knowledge, BMP policies, BMP costs, BMP

benefits, and BMP cost effectiveness. To address the challenge, the aim of this research is to

develop a WebGIS-based decision support system to facilitate information communications for

agricultural BMP adoption.

The Chapter 2 reviewed literature in the related research domains and identified research

gaps. Specifically, BMP adoption studies were analyzed, and information needs of stakeholders

138

for BMP adoption were identified. Many modelling systems have been developed to generate

valuable information for BMP evaluation (Srivastava et al., 2002; Yang et al., 2003; Turpin et

al., 2005). However, those modelling systems and desktop-based interfaces were commonly too

complex for farmers and conservation managers to operate. Moreover, despite the fact that some

applications of GIS have been developed to facilitate watershed modelling and BMP evaluation

(Jayakrishnan et al., 2005; Olivera et al., 2006; Liu, Bralts, & Engel, 2015; Karki et al., 2017;

Jang, Ahn, & Kim, 2017), an identified need is to extend those GIS systems to improve

stakeholders’ access to information and facilitate their communications to address the possible

adoption concerns.

The Chapter 3 developed an information model to characterize information content,

communications, and related technologies and tools in the information communication process

for BMP adoption. The development of the information model was based on the analysis of

information needs and the use of ICTs. The information needs of farmers and conservation

managers were classified into two categories: public information and BMP planning information.

The characteristics of different ICTs and the potential of ICTs were discussed to understand the

use of ICTs to support various information needs of stakeholders.

The Chapter 4 introduced the design of a WebGIS-based decision support system for

facilitating agricultural BMP adoption. The task-oriented design approach was used to

implement the information model and guide the design of the WebGIS-based decision support

system at three design levels: subsystem, module, and component level. Subsystems were

designed to support the information needs of stakeholders, modules of each subsystem were

designed to support information tasks required to achieve the information needs in the

139

information communication process, and components at the client, service and data tiers of

modules were designed to support user interactions with the interface to complete information

tasks.

The Chapter 5 showcased a prototype of WebGIS-based decision support system, which

was adapted to the Gully Creek watershed. The procedure of system localization for the Gully

Creek watershed was discussed, which can provide references for transferring the WebGIS-based

decision support system to other watersheds. In the demonstration of the prototype, the user

interfaces (interface layout and components) were introduced, user interactions with the interface

were described, and the means of information presentations were illustrated. The demonstration

showed how the information needs of stakeholders were met and information communications

among stakeholders were fulfilled through the WebGIS-based decision support system.

The chapter 6 evaluated the WebGIS-based decision support system through direct use

and demonstration. The evaluation by direct use identified violence to usability principles, while

the evaluation during demonstration evaluated user task and information of the system. The

results from evaluation by direct use and evaluation during demonstration were aggregated for

assessment of the system usability. The evaluation results revealed that the evaluators overall

satisfied with the system. Specifically, the system interface was easy to use, the web services

were responsive, and the functions and information satisfied their task requirements for

agricultural BMP adoption.

140

7.2 Research contributions

Theoretical contribution

This research examined BMP adoption as an information communication process

wherein stakeholders request information to improve their knowledge and collaboratively

address their concerns towards BMP adoption. Based on the innovation diffusion theory (Rogers,

1995), the research identified key stakeholders and defined the role of stakeholders in the

information communication process. Specifically, scientific researchers, conservation managers

and farmers are the key stakeholders for BMP adoption. Scientific researchers play an important

role as they identify environmental problems, design BMPs and generate the information on

economic costs, environmental benefits and cost-effectiveness of BMPs. Farmers are adopters of

BMPs. Conservation managers are BMP policy designers and facilitators for BMP adoption.

During the process, both farmers and conservation managers need information to improve their

understanding on BMPs and their effects (Shao et al., 2017). They also need have frequent

communications on the information on economic costs, environmental benefits, and cost-

effectiveness of BMPs to reach consensus on BMP adoption decisions.

In addition to a variety of social, cultural and economic factors, information is an

essential factor for the adoption process of agricultural BMPs (Rogers, 1995; Alonge and Martin,

1995; Kaiser et al., 1999; Smithers & Furman, 2003; Dietz et al.2004; Claassen et al., 2008;

Atari et al., 2009). Based on the understanding on agricultural BMP adoption process, the

research developed an information model to conceptualize the information processes for BMP

adoption. The information model improves knowledge on how the information communication

process for agricultural BMP adoption can be achieved by the application of GIS, integrated

141

economic-hydrologic models and ICTs. Specifically, a WebGIS can be used to facilitate

information sharing among farmers about their field characteristics, environmental concerns, and

BMP adoption (Kingston et al., 2000). The public information portal can be used to disseminate

environmental policies and BMP related technical knowledge (Janssen & Kies, 2005). The

WebGIS, coupled with an economic-hydrologic model and an optimization model, can improve

stakeholders’ access to the BMP planning information (Goodchild et al., 1993). Moreover, the

communication tools such as discussion forum and email can be used facilitate communications

among stakeholders (Sidlar and Rinner, 2007; Butt and Li, 2012).

Methodological contribution

Existing GIS applications for BMP evaluation were mostly used by experts (Jayakrishnan

et al., 2005; Olivera et al., 2006; Liu, Bralts, & Engel, 2015; Karki et al., 2017; Jang, Ahn, &

Kim, 2017). The innovation of the WebGIS-based decision support system is to communicate

BMP evaluation modelling information to stakeholders including farmers and conservation

managers. This research used a task-oriented design approach to design a user-friendly WebGIS-

based decision support system for facilitating the adoption of agricultural BMPs. Based on the

understanding on the information communication process presented by the information model, a

hierarchy of user tasks was developed to facilitate the achievement of the information model and

guide the system design at three different design levels: the subsystem, module and component

levels.

The design of the WebGIS-based decision support system extends the desktop-based

integrated economic-hydrologic modelling systems for BMP adoption by allowing stakeholders

to easily access a wider range of information which includes public information (field

142

characteristics, environmental concerns, BMP adoption status, BMP technical knowledge, BMP

policies) and BMP planning information (BMP costs, BMP benefits, BMP cost effectiveness,

and spatial targeting of BMPs). The WebGIS-based decision support system also extends the

traditional desktop GIS by offering a collaborative environment for stakeholders and improving

information communications between farmers and conservation managers to reach consensus on

BMP adoption.

Practical contribution

This research developed a prototype of the WebGIS-based decision support system for

the Gully Creek watershed. The software for the prototype development was discussed. The

localization of the design to the Gully Creek watershed was outlined. By adapting the system

design to the Gully Creek watershed, the WebGIS-based decision support system prototype

demonstrates that the system has the potential to be transferred to other watersheds with various

customization tasks.

The system prototype supports the information communication process for BMP adoption

and facilitates farmers and conservation managers in the Gully Creek watershed to conduct

various tasks related to BMP adoption. In particular, the system allows farmers and conservation

managers to evaluate economic costs, environmental benefits, and cost-effectiveness of four

BMPs (i.e. conservation tillage, cover crop, nutrient management and WASCoBs) on farm fields.

The system also supports the conservation manager to design economic and environmental

policies for BMP implementation in the Gully Creek watershed.

143

7.3 Future study

The WebGIS-based decision support system has the potential to be further developed to

support agricultural BMP adoption in various watershed contexts:

1) Scenario scripts were utilized as a tool to derive key user tasks and design

requirements based on an existing watershed BMP modelling project in the research group with

inputs from staff members at conservation authority and government. These scenario scripts can

be further developed by interviewing representative farmers and conservation managers from

other conservation organizations.

2) The architecture of the WebGIS-based decision support system included key system

components for supporting agricultural BMP adoption. However, the system design can be

further modified and/or expanded to include more information communication tasks. For

example, a module can be designed to show the historical results of BMP adoption. Moreover, a

community forum can be incorporated into the system to enhance the information sharing among

a wide range of stakeholders to further facilitate the adoption of BMPs.

3) The system can be further improved to support different parameterization schemes,

such as climate scenarios. In this regard, enhancements of system should be made. Especially, an

interface between the WebGIS interface and the integrated economic-hydrologic model can be

built to allow users to modify model inputs or parameters. Instead of arbitrarily assigning fixed

values of parameters, this interface should enable the model to explore a parameter space to

generate a range of BMP cost-effectiveness. By evaluating BMPs subject to the parameter space,

144

a range of BMP cost effectiveness can be obtained, and the uncertainty or confidence interval of

BMP cost effectiveness can be calculated.

4) A prototype of the WebGIS-based decision support system for the Gully Creek

watershed was developed based on four representative BMPs which included conservation

tillage, cover crop, nutrient management, and WASCoBs. The engine of the system was based on

existing farm economic, watershed hydrologic, and integrated modelling for the watershed.

Various workflows can be developed for adapting the system to other watersheds with various

BMPs, which improve the transferability of the WebGIS-based decision support system. The

workflows may include basic dataset collection, modelling development, and system function

development.

5) The prototype of the WebGIS-based decision support system for the Gully Creek

watershed was evaluated through direct use and demonstration. Based on the evaluation results, a

very positive feedback was received. However, local farmers were not included in the evaluation

due to the sensitiveness in communicating with farmers by people outside conservation authority

and government. Further work may include working with conservation authority and government

to develop a trust relationship with local farmers and conduct a user evaluation by them to

understand the strengths and limitations of the system.

6) This study designed the WebGIS-based decision support system to facilitate

information communication among stakeholders for facilitating agricultural BMP adoption.

Future studies can be developed to quantify the effects of the WebGIS-based decision support

system on improving BMP adoption. Specifically, efforts can be made to understand the

145

potential usage of the WebGIS-based decision support system among stakeholders, and to what

extent the system usage contributes to BMP adoption.

146

REFERENCES

Ackerman, D., & Stein, E. D. (2008). Evaluating the Effectiveness of Best Management

Practices Using Dynamic Modeling. Journal of Environmental Engineering, 134(8).

https://doi.org/10.1061/(ASCE)0733-9372(2008)134:8(628)

Ahmed, A.A.N., Haque, H.F., Rahman, A., Ashraf, M.S., Saha, S., Shatabda, S. (2017). A

Participatory Sensing Framework for Environment Pollution Monitoring and Management.

Retrieved from: https://arxiv.org/abs/1701.06429

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision

Processes, 50(2), 179-211. https://doi.org/10.1016/0749-5978(91)90020-T

Alonge, A. J., & Martin, R. A. (1995). Assessment of the adoption of sustainable agriculture

practices: Implications for agricultural education. Journal of Agricultural Education, 36(3),

34-42.

Annett, J. (2004). Hierarchical task analysis. In N. Stanton, A. Hedge, K. Brookhuis, E. Salas, &

H. Hendrick (Eds.), Handbook of Human Factors and Ergonomics Methods. Boca Raton,

FL: CRC Press.

Arabi, M., Frankenberger, J. R., Engel, B. A., & Arnold, J. G. (2008). Representation of

agricultural conservation practices with SWAT. Hydrological Processes, 22(16), 3042-

3055. https://doi.org/10.1002/hyp.6890

147

Arnold, J. G., Srinivasan, R., Muttiah, R. S., & Williams, J. R. (1998). Large area hydrologic

modeling and assessment part I: Model development. Journal of the American Water

Resources Association, 34(1), 73-89. https://doi.org/10.1111/j.1752-1688.1998.tb05961.x

Atari, D. O. A., Yiridoe, E. K., Smale, S., & Duinker, P. N. (2009). What motivates farmers to

participate in the Nova Scotia environmental farm plan program? Evidence and

environmental policy implications. Journal of Environmental Management, 90(2), 1269-

1279. https://doi.org/10.1016/j.jenvman.2008.07.006

Barkmann, H., Lincke, R., & Löwe, W. (2009). Quantitative evaluation of software quality

metrics in open-source projects. In Proceedings - International Conference on Advanced

Information Networking and Applications, Bradford (pp.1067-1072).

https://doi.org/10.1109/WAINA.2009.190

Baumgart-Getz, A., Prokopy, L. S., & Floress, K. (2012). Why farmers adopt best management

practice in the United States: A meta-analysis of the adoption literature. Journal of

Environmental Management, 96(1), 17-25. https://doi.org/10.1016/j.jenvman.2011.10.006

Bicknell, B. R., Imhoff, J. C., Kittle, Jr., J. L., Jobes, T. H., & Donigian, Jr., A. S. (2005).

Hydrological simulation program - Fortran: HSPF version 12.2 user’s manual. Cooperation

with the US Geological Survey and US Environmental Protection Agency

Bingner, R. L., & Theurer, F. D. (2003). AnnAGNPS technical processes: documentation version

3.2. Retrieved from

https://www.ars.usda.gov/ARSUserFiles/60600505/AGNPS/PLModel/Document/Tech_Do

c.PDF

148

Bootstrap vs. Foundation (2019), Retrieved from https://stackshare.io/stackups/bootstrap-vs-

foundation

Boroushaki, S., & Malczewski, J. (2010a). Using the fuzzy majority approach for GIS-based

multicriteria group decision-making. Computers and Geosciences, 36(3), 302-312.

https://doi.org/10.1016/j.cageo.2009.05.011

Boroushaki, S., & Malczewski, J. (2010b). Measuring consensus for collaborative decision-

making: A GIS-based approach. Computers, Environment and Urban Systems, 34(4), 322-

332. https://doi.org/10.1016/j.compenvurbsys.2010.02.006

Böhlke, J. K. (2002). Groundwater recharge and agricultural contamination. Hydrogeology

Journal, 10(1), 153-179. https://doi.org/10.1007/s10040-001-0183-3

Breetz, H. L., Fisher-Vanden, K., Jacobs, H., & Schary, C. (2005). Trust and communication:

mechanisms for increasing farmers’ participation in water quality trading. Land Economics,

81(2), 170-190. https://doi.org/10.3368/le.81.2.170

Brent Hall, G., Chipeniuk, R., Feick, R. D., Leahy, M. G., & Deparday, V. (2010). Community-

based production of geographic information using open source software and Web 2.0.

International Journal of Geographical Information Science, 24(5), 761-781.

https://doi.org/10.1080/13658810903213288

Brown, G., & Kyttä, M. (2014). Key issues and research priorities for public participation GIS

(PPGIS): A synthesis based on empirical research. Applied Geography, 46, 122-136.

https://doi.org/10.1016/j.apgeog.2013.11.004

149

Brown, G. (2012). Public Participation GIS (PPGIS) for regional and environmental planning:

feflections on a decade of empirical research. Journal of the Urban and Regional

Information Systems Association (URISA), 25(2), 7-18. https://doi.org/10.1068/b2751t

Bugs, G., Granell, C., Fonts, O., Huerta, J., & Painho, M. (2010). An assessment of Public

Participation GIS and Web 2.0 technologies in urban planning practice in Canela, Brazil.

Cities, 27(3), 172-181. https://doi.org/10.1016/j.cities.2009.11.008

Bultena, G. L., & Hoiberg, E. O. (1983). Factors affecting farmer adoption of conservation

tillage. Journal of Soil and Water Conservation, 38(3), 281-284.

Butt, M. A., & Li, S. (2012). Open source based online map sharing to support real-time

collaboration. OSGeo Journal, 10, 4-15.

Butt, M. A., & Li, S. (2012). Developing a web-based, collaborative PPGIS prototype to support

public participation. Applied Geomatics, 4(3), 197-215. https://doi.org/10.1007/s12518-

012-0085-1

Carlson, J. E., D. A. Dillman and W. R. Lassey. (1981). The farmer and erosion: Factors

influencing the use of control practices. Bulletin No. 601, Agricultural Experiment Station,

University of Idaho, Moscow, Idaho.

Carroll, J. M. (1997). Human-computer interaction: Psychology as a science of design.

International Journal of Human Computer Studies, 48, 61-83.

https://doi.org/10.1006/ijhc.1996.0101

150

Chapman, R., Slaymaker, T., & Paper, W. (2002). ICTs and Rural Development: Review of the

Literature, Current Interventions and Opportunities for Action. Retrieved from

https://www.odi.org/resources/docs/2670.pdf

Chin, J. P., Diehl, V. a and Norman, L. K. (1988) ‘Development of an instrument measuring user

satisfaction of the human-computer interface’, CHI ’88- Proceedings of the SIGCHI

conference on Human factors in computing systems, pp. 213–218. doi:

10.1145/57167.57203.

Claassen, R., Cattaneo, A., & Johansson, R. (2008). Cost-effective design of agri-environmental

payment programs: U.S. experience in theory and practice. Ecological Economics, 65(4),

737-752. https://doi.org/10.1016/j.ecolecon.2007.07.032

Coggan, A., Whitten, S. M., & Bennett, J. (2010). Influences of transaction costs in

environmental policy. Ecological Economics, 69(9), 1777-1784.

https://doi.org/10.1016/j.ecolecon.2010.04.015

Coleman, D. J., Georgiadou, Y., Labonte, J., Observation, E., & Canada, N. R. (2009).

Volunteered Geographic Information: the nature and motivation of produsers.

GeoInformation Science, 4, 332-358.

Dabney, S. M. (1998). Cover crop impacts on watershed hydrology. Journal of Soil and Water

Conservation, 53(3), 207-213.

151

DeLone, W. H. & Mclean, E. R. (2003). The DeLone and McLean Model of Information

Systems Success: A Ten-Year Update. Journal of Management Information Systems /

Spring, 19(4), 9–30. doi: 10.1073/pnas.0914199107.

D’souza, G., Cyphers, D., & Phipps, T. (1993). Factors affecting the adoption of sustainable

Agricultural Practices. Agricultural and Resource Economics Review, 22(2), 159-165.

https://doi.org/10.1017/S1068280500004743

Dietz, M. E., Clausen, J. C., & Filchak, K. K. (2004). Education and changes in residential

nonpoint source pollution. Environmental Management, 34(5), 684-690.

https://doi.org/10.1007/s00267-003-0238-4

Dupont, D. P. (2010). Cost-sharing incentive programs for source water protection: The Grand

River’s rural water quality program. Canadian Journal of Agricultural Economics, 58(4),

481-496. https://doi.org/10.1111/j.1744-7976.2010.01197.x

Ebert, P., & Shapiro, S. (2009). The good, the bad and the ugly. Synthese, 170(3), 415-441.

https://doi.org/10.1007/s11229-007-9263-9

Elwood, S., & Ghose, R. (2011). PPGIS in community development planning: Framing the

organizational context. In Classics in Cartography: Reflections on Influential Articles from

Cartographica (pp:83-105). New York, NY: John Wiley & Son Ltd.

https://doi.org/10.1002/9780470669488.ch8

Elwood, S., Goodchild, M. F., & Sui, D. Z. (2012). Researching Volunteered Geographic

Information: Spatial Data, Geographic Research, and New Social Practice, 102(3), 571-

152

590. Annals of the Association of American Geographers.

https://doi.org/10.1080/00045608.2011.595657

Engel, S., Pagiola, S., & Wunder, S. (2008). Designing payments for environmental services in

theory and practice: An overview of the issues. Ecological Economics, 65(4), 663-674.

https://doi.org/10.1016/j.ecolecon.2008.03.011

Environmental Protection Agency (US EPA) 1999, Preliminary Data Summary of Urban

Stormwater Best Management Practices, United States Environmental Protection Agency

report No. EPA- 821-R-99-012: www.epa.gov/waterscience/stormwater

Eppler, M. J., & Muenzenmayer, P. (2002). Measuring Information Quality in the Web Context:

A survey of state-of-art instruments and an application methodology (Practice-Oriented).

Proceedings of the Seventh International Conference on Information Quality (ICIQ-02) (pp.

187-196). https://doi.org/10.1.1.477.4680

Falconer, K. (2000). Farm-level constraints on agri-environmental scheme participation: A

transactional perspective. Journal of Rural Studies, 16(3), 379-394.

https://doi.org/10.1016/S0743-0167(99)00066-2

Feather, P. M., & Amacher, G. S. (1994). Role of information in the adoption of best

management practices for water quality improvement. Agricultural Economics, 11, 159-

170.

Fischer, G. (1993). Beyond human computer interaction: Designing useful and usable

computational environments. In: People and Computers VIII: Proceedings of the HCI'93

153

Conference (Loughborough, England), Cambridge University Press, Cambridge, UK, pp.

17-31.

Fischer, G. (2001). User modeling in human-computer interaction. User Modeling and User-

Adapted Interaction, 11, 65-86. https://doi.org/10.1023/A:1011145532042

Flanagin, A. J., & Metzger, M. J. (2008). The credibility of volunteered geographic

information. GeoJournal, 72, 137-148. https://doi.org/10.1007/s10708-008-9188-y

Fossey, E., Harvey, C., McDermott, F., & Davidson, L. (2002). Understanding and evaluating

qualitative research. Australian and New Zealand Journal of Psychiatry, 36, 717-732.

https://doi.org/10.1046/j.1440-1614.2002.01100.x

Garbarino, S., & Holland, J. (2009). Quantitative and qualitative methods in impact evaluation

and measuring results. Governance and Social Development Resource Center. Discussion

Paper. Birmingham: University of Birmingham.

Garnett, T., Appleby, M. C., Balmford, A., Bateman, I. J., Benton, T. G., Bloomer, P., …

Godfray, H. C. J. (2013). Sustainable intensification in agriculture: Premises and policies.

Science, 341(6141), 33-34. https://doi.org/10.1126/science.1234485

Ghebremichael, L. T., Veith, T. L., & Hamlett, J. M. (2013). Integrated watershed- and farm-

scale modeling framework for targeting critical source areas while maintaining farm

economic viability. Journal of Environmental Management, 114, 381-394.

https://doi.org/10.1016/j.jenvman.2012.10.034

154

Gitau, M. W., Veith, T. L., & Gburek, W. J. (2004). Farm-level optimization of BMP placement

for cost-efficient pollution reduction. Transactions of the ASAE, 47(6), 1923-1931.

https://doi.org/10.13031/2013.17805

Glendenning, C. J., & Ficarelli, P. P. (2012). The relevance of content in ICT initiatives in Indian

agriculture. International Food Policy Research Institute. Retrieved from

http://www.researchintouse.com/resources/riu11discuss16info-comms.pdf

GLRI funding. (n, d), Retrieved from https://www.glri.us/funding

Goodchild, M. F., Parks, B. O., & Steyaert, L. T. (1993). Environmental modeling with GIS.

New York: Oxford University Press

Goodchild, M. F., & Li, L. (2012). Assuring the quality of volunteered geographic information.

Spatial Statistics, 1, 110-120. https://doi.org/10.1016/j.spasta.2012.03.002

Gratier, T., Spencer, P., Hazzard, E. (2015): OpenLayers 3: Beginners's Guide, Packt Publishing Ltd.,

Birmingham, UK

Greiner, R., & Gregg, D. (2011). Farmers’ intrinsic motivations, barriers to the adoption of

conservation practices and effectiveness of policy instruments: Empirical evidence from

northern Australia. Land Use Policy, 28(1), 257-265.

https://doi.org/10.1016/j.landusepol.2010.06.006

Grinnell, R. M., Jr. (Ed.). (2001). Social work research and evaluation: Quantitative and

qualitative approaches (7th ed.). Belmont, CA: Wadsworth.

155

Hamilton, S., & Chervany, N. L. (2006). Evaluating Information System Effectiveness - Part I:

Comparing Evaluation Approaches. MIS Quarterly, 5(3), 55-69.

https://doi.org/10.2307/249291

Havlin, J. L., Beaton, J. D., Tisdale, S. L., & Nelson, W. L. (1999). Soil Fertility and Fertilizers:

An Introduction to Nutrient Management. In Soil Fertility and Fertilizers. NYC, NY:

Pearson https://doi.org/10.1007/s10310-009-0124-x

Holzinger, a (2005) ‘Usability engineering methods for software developers’, Communications

of the ACM, 48(1), pp. 71–74. doi: 10.1145/1039539.1039541.

Issa T., & Isaias P. (2015) Usability and Human Computer Interaction (HCI) (pp. 19-36). In:

Sustainable Design. Springer, London

Jang, S. S., Ahn, S. R., & Kim, S. J. (2017). Evaluation of executable best management practices

in Haean highland agricultural catchment of South Korea using SWAT. Agricultural Water

Management, 180(Part B), 224-234. https://doi.org/10.1016/j.agwat.2016.06.008

Janssen, D., & Kies, R. (2005). Online forums and deliberative democracy. Acta Politica, 40(3),

317-335. https://doi.org/10.1057/palgrave.ap.5500115

Jayakrishnan, R., Srinivasan, R., Santhi, C., & Arnold, J. G. (2005). Advances in the application

of the SWAT model for water resources management. Hydrological Processes, 19(3), 749-

762. https://doi.org/10.1002/hyp.5624

156

Jeffries, R. et al. (1991) ‘User interface evaluation in the real world: a comparison of four

techniques’, in Proceedings of the SIGCHI conference on Human factors in computing

systems, pp. 119–124. doi: 10.1145/108844.108862.

Jeffries, R. and Desurvire, H. (1992) ‘Usability testing vs. heuristic evaluation’, ACM SIGCHI

Bulletin, 24(4), pp. 39–41. doi: 10.1145/142167.142179.

Kaiser, F. G., Wölfing, S., & Fuhrer, U. (1999). Environmental attitude and ecological

behaviour. Journal of Environmental Psychology, 19(1), 1-19.

https://doi.org/10.1006/jevp.1998.0107

Kaplan, B., & Maxwell, J. A. (2006). Qualitative Research Methods for Evaluating Computer

Information Systems. In Evaluating the Organizational Impact of Healthcare Information

Systems,30-55. https://doi.org/10.1007/0-387-30329-4_2

Karatzas K., Peinel G., Rose T. (2000): GIS-based Dissemination of Air Quality Information in

APNEE. 6th European Commission GI and GIS Workshop, The Spatial Information

Society - Shaping the Future, Lyon, France.

Karki, R., Tagert, M. L. M., Paz, J. O., & Bingner, R. L. (2017). Application of AnnAGNPS to

model an agricultural watershed in East-Central Mississippi for the evaluation of an on-farm

water storage (OFWS) system. Agricultural Water Management, 192, 103-114.

https://doi.org/10.1016/j.agwat.2017.07.002

157

Kelly, K. E., Belcher, K., & Khakbazan, M. (2018). Economic targeting of agricultural beneficial

management practices to address phosphorus runoff in Manitoba. Canadian Journal of

Agricultural Economics, 66(1), 143-166. https://doi.org/10.1111/cjag.12143

Kershner, R., Mercer, N., Warwick, P., & Staarman, J. K. (2010). Can the interactive whiteboard

support young children’s collaborative communication and thinking in classroom science

activities? International Journal of Computer-Supported Collaborative Learning, 5(4), 359-

383. https://doi.org/10.1007/s11412-010-9096-2

Keßler, C. 2004. Design and implementation of argumentation maps. Münster, Germany:

Diploma Thesis, Westfälische Wilhelms-Universitäty Münster, Germany. Available online

at http://www.carstenkessler.de/argumap/.

Kingston, R., Carver, S., Evans, A., & Turton, I. (2000). Web-based public participation

geographical information systems: An aid to local environmental decision-making.

Computers, Environment and Urban Systems, 24(2), 109-125.

https://doi.org/10.1016/S0198-9715(99)00049-6

Kingston, R. (2007). Public Participation in Local Policy Decision-making: The Role of Web-

based Mapping. The Cartographic Journal, 44(2), 138-144.

https://doi.org/10.1179/000870407X213459

Kitajima, M. (2018). Cognitive Walkthrough for the Web. Journal of Japan Society for Fuzzy

Theory and Systems. https://doi.org/10.3156/jfuzzy.14.5_14

158

Knowler, D., & Bradshaw, B. (2007). Farmers’ adoption of conservation agriculture: A review

and synthesis of recent research. Food Policy, 32(1), 25-48.

https://doi.org/10.1016/j.foodpol.2006.01.003

Leach, W. D., Pelkey, N. W., & Sabatier, P. A. (2002). Stakeholder partnerships as collaborative

policymaking: Evaluation criteria applied to watershed management in California and

Washington. Journal of Policy Analysis and Management, 21(4), 645-670.

https://doi.org/10.1002/pam.10079

Lee, T., Rister, M. E., Narashimhan, B., Srinivasan, R., Andrew, D., & Ernst, M. R. (2010).

Evaluation and spatially distributed analyses of proposed cost-effective BMPs for reducing

phosphorus level in Cedar Creek Reservoir, Texas. Transactions of the ASABE, 53(5),

1619-1627.

Lee, Y. W. et al. (2002). AIMQ: A methodology for information quality assessment. Information

and Management, 40, 133-146. https://doi.org/10.1016/S0378-7206(02)00043-5

Leung, L. (2015). Validity, reliability, and generalizability in qualitative research. Journal of

Family Medicine and Primary Care, 4(3), 324–327. https://doi.org/10.4103/2249-

4863.161306

Lewis, C., & Rieman, J. (1994). Task-Centered User Interface Design: A Practical Introduction

[PDF file]. Retrieved from ftp.cs.colorado.edu/pub/cs/distribs/clewis/HCI-Design- Book.

Liu, Y., Bralts, V. F., & Engel, B. A. (2015). Evaluating the effectiveness of management

practices on hydrology and water quality at watershed scale with a rainfall-runoff model.

159

Science of the Total Environment, 511, 298-308.

https://doi.org/10.1016/j.scitotenv.2014.12.077

Llewellyn, R. S. (2007). Information quality and effectiveness for more rapid adoption decisions

by farmers. Field Crops Research, 104(1-3), 148-156.

https://doi.org/10.1016/j.fcr.2007.03.022

Logan, T. J. (1993). Agricultural best management practices for water pollution control: current

issues. Agriculture, Ecosystems and Environment, 46(1-4), 223–231.

https://doi.org/10.1016/0167-8809(93)90026-L

Lu, Y., Song, S., Wang, R., Liu, Z., Meng, J., Sweetman, A. J., … Wang, T. (2015). Impacts of

soil and water pollution on food safety and health risks in China. Environment International,

77, 5-15. https://doi.org/10.1016/j.envint.2014.12.010

Lubell, M., & Fulton, A. (2007). Local diffusion networks act as pathways to sustainable

agriculture in the Sacramento River Valley. California Agriculture, 61(3), 131-137.

https://doi.org/10.3733/ca.v061n03p131

Luchette, J., & Crawford, T. (2008). A public participation GIS application for citizen-based

watershed monitoring in the Pamlico-Tar River basin, North Carolina. Southeastern

Geographer, 48(2), 184-200 https://doi.org/10.1353/sgo.0.0022

Luzar, E. J., & Diagne, A. (1999). Participation in the next generation of agriculture conservation

programs: The role of environmental attitudes. Journal of Socio-Economics, 28(3), 335-349.

https://doi.org/10.1016/S1053-5357(99)00021-9

160

Ma, S., Swinton, S. M., Lupi, F., & Jolejole-Foreman, C. (2012). Farmers’ Willingness to

Participate in Payment-for-Environmental-Services Programmes. Journal of Agricultural

Economics, 63(3), 604-626. https://doi.org/10.1111/j.1477-9552.2012.00358.x

Maharjan, G. R., Ruidisch, M., Shope, C. L., Choi, K., Huwe, B., Kim, S. J., … Arnhold, S.

(2016). Assessing the effectiveness of split fertilization and cover crop cultivation in order

to conserve soil and water resources and improve crop productivity. Agricultural Water

Management, 163(1), 305-318. https://doi.org/10.1016/j.agwat.2015.10.005

Malczewski, J. (2004). GIS-based land-use suitability analysis: A critical overview. Progress in

Planning, 62(1), 3-65. https://doi.org/10.1016/j.progress.2003.09.002

Mankoff, J. et al. (2003) ‘Heuristic evaluation of ambient displays’, in Proceedings of the

conference on Human factors in computing systems - CHI ’03, p. 169. doi:

10.1145/642611.642642.

Matson, P. A., Parton, W. J., Power, A. G., & Swift, M. J. (1997). Agricultural intensification

and ecosystem properties. Science, 277(5325), 504-509.

https://doi.org/10.1126/science.277.5325.504

Mazza, R., & A. Berre. (2007). Focus Group Methodology for Evaluating Information

Visualization Techniques and Tools. In 11th International Conference of Information

Visualization. Zürich, Switzerland, (pp. 74-80).

161

Mbilinyi, B. P., Tumbo, S. D., Mahoo, H. F., & Mkiramwinyi, F. O. (2007). GIS-based decision

support system for identifying potential sites for rainwater harvesting. Physics and

Chemistry of the Earth, 32(15-18), 1074-1081. https://doi.org/10.1016/j.pce.2007.07.014

McCown, R. L. (2002). Locating agricultural decision support systems in the troubled past and

socio-technical complexity of “models for management.” Agricultural Systems, 74(1), 11-

25. https://doi.org/10.1016/S0308-521X(02)00020-3

McCoy, N., Chao, B., & Gang, D. D. (2015). Nonpoint source pollution (article). Water

Environment Research, 87(10), 1576–1594. https://doi.org/10.2175/106143015

x14338845156263

Morrison, K., & FitzGibbon, J. E. (2014). Adaptive governance of dynamic social-ecological

systems: the case of the ontario Environmental Farm Plan (1992-2011). Agroecology and

Sustainable Food Systems, 38(4), 378-409. https://doi.org/10.1080/21683565.2013.870627

Murray, M. (2002). An investigation of specifications for migrating to a web portal framework

for the dissemination of health information within a public health network. Paper presented

at Proceedings of the 35th Annual Hawaii International Conference on System Science, Big

Island, Hawaii, USA (pp. 1917-1925). USA: IEEE Computer Society, Los Alamitos,

California, USA

Nielsen, J. (1990) ‘Ten Usability Heuristics’, Communications of the ACM, 3(1990), pp. 1–2.

Available at: http://www.useit.com/papers/heuristic/heuristic_list.html.

Nielsen, J. (1993) Usability Engineering, Usability Engineering. doi: 10.1145/1508044.1508050.

162

Nielsen, J. (1994). Usability inspection methods. In Conference companion on Human factors in

computing systems - CHI ’94. https://doi.org/10.1145/259963.260531

Nxumalo, K. K. S., & Oladele, O. I. (2013). Factors affecting farmers’ participation in

agricultural programme in Zululand District, Kwazulu Natal Province, South Africa.

Journal of Social Sciences, 34(1), 83-88. https://doi.org/10.1080/09718923.2013.11893120

Olivera, F., Valenzuela, M., Srinivasan, R., Choi, J., Cho, H., Koka, S., & Agrawal, A. (2006).

ArcGIS-SWAT: A geodata model and GIS interface for SWAT. Journal of the American

Water Resources Association, 42(2), 295-309. https://doi.org/10.1111/j.1752-

1688.2006.tb03839.x

Open maps. (n.d.) Retrieved from https://open.canada.ca/en/open-maps.

Osei, E., Gassmann, P. W., Jones, R. D., Pratt, S. J., Hauck, L. M., Beran, L. J., … Williams, J.

R. (2000). Economic and environmental impacts of alternative practices on dairy farms in

an agricultural watershed. Journal of Soil and Water Conservation, 55(4), 466-472.

Palmer, J. W. (2002). Web site usability, design, and performance metrics. Information Systems

Research, 13(2), 151–167. doi: 10.1287/isre.13.2.151.88.

Panteli, N., & Dawson, P. (2001). Video conferencing meetings: Changing patterns of business

communication. New Technology, Work and Employment, 16(2), 88-99.

https://doi.org/10.1111/1468-005X.00079

163

Parajuli, P. B., Mankin, K. R., & Barnes, P. L. (2008). Applicability of targeting vegetative filter

strips to abate fecal bacteria and sediment yield using SWAT. Agricultural Water

Management, 95(10), 1189-1200. https://doi.org/10.1016/j.agwat.2008.05.006

Peris, R., Gimeno, M. A., Pinazo, D., Ortet, G., Carrero, V., Sanchiz, M., & Ibáñez, I. (2002).

Online chat rooms: Virtual spaces of interaction for socially oriented people.

CyberPsychology & Behavior, 5(1), 43-51. https://doi.org/10.1089/109493102753685872

Pinelle, D., Wong, N. and Stach, T. (2008) ‘Heuristic evaluation for games: usability principles

for video game design’, Proceedings of SIGCHI Conference on Human Factors in

Computing Systems, pp. 1453–1462. doi: 10.1145/1357054.1357282.

Plummer, R., Spiers, A., Summer, R., & FitzGibbon, J. (2008). The contributions of stewardship

to managing agro-ecosystem environments. Journal of Sustainable Agriculture, 31(3), 55-

84. https://doi.org/10.1300/J064v31n03_06

Prager, K., & Freese, J. (2009). Stakeholder involvement in agri-environmental policy making -

Learning from a local- and a state-level approach in Germany. Journal of Environmental

Management, 90(2), 1154-1167. https://doi.org/10.1016/j.jenvman.2008.05.005

Prager, K., Reed, M., & Scott, A. (2012). Encouraging collaboration for the provision of

ecosystem services at a landscape scale-Rethinking agri-environmental payments. Land Use

Policy, 29(1), 244-249. https://doi.org/10.1016/j.landusepol.2011.06.012

Pricing for maps, routes, and places (2019), Retrieved from https://cloud.google.com/maps-

platform/pricing/sheet/

164

Prokopy, L. S., Floress, K., Klotthor-Weinkauf, D., & Baumgart-Getz, A. (2008). Determinants

of agricultural best management practice adoption: Evidence from the literature. Journal of

Soil and Water Conservation, 63(5), 300-311. https://doi.org/10.2489/jswc.63.5.300

Qi, H., & Altinakar, M. S. (2011). A GIS-based decision support system for integrated flood

management under uncertainty with two dimensional numerical simulations. Environmental

Modelling and Software, 26(6), 817-821. https://doi.org/10.1016/j.envsoft.2010.11.006

Qi, H., & Altinakar, M. S. (2011). A conceptual framework of agricultural land use planning

with BMP for integrated watershed management. Journal of Environmental Management,

92(1), 149-155. https://doi.org/10.1016/j.jenvman.2010.08.023

Rao, M., Fan, G., Thomas, J., Cherian, G., Chudiwale, V., & Awawdeh, M. (2007). A web-based

GIS Decision Support System for managing and planning USDA’s Conservation Reserve

Program (CRP). Environmental Modelling and Software, 22(9), 1270-1280.

https://doi.org/10.1016/j.envsoft.2006.08.003

Reimer, A. P., Thompson, A. W., & Prokopy, L. S. (2012). The multi-dimensional nature of

environmental attitudes among farmers in Indiana: Implications for conservation adoption.

Agriculture and Human Values, 29(1), 29-40. https://doi.org/10.1007/s10460-011-9308-z

Reimer, A. P., Weinkauf, D. K., & Prokopy, L. S. (2012). The influence of perceptions of

practice characteristics: An examination of agricultural best management practice adoption

in two indiana watersheds. Journal of Rural Studies, 28(1), 118-128.

https://doi.org/10.1016/j.jrurstud.2011.09.005

165

Richardson, D. (2006). How can agricultural extension best harness ICTs to improve rural

livelihood in developing countries? In ICT in Agriculture: Perspectives of Technological

Innovation (ed) E. Gelb, A. European Federation for Information Technologies in

Agriculture, Food and the Environment p. 14

Rinner, C. (2001). Argumentation maps: GIS-based discussion support for on-line planning.

Environment and Planning B: Planning and Design, 28(6), 847-863.

https://doi.org/10.1068/b2748t

Rogers, E. M. (1995). Diffusion of Innovations, Fourth Edition, the Free Press, New York

Schlossberg, M., & Shuford, E. (2005). Delineating “public” and “participation” in PPGIS.

URISA Journal, 16, 15-26.

Shao, H., Yang, W., Lindsay, J., Liu, Y., Yu, Z., & Oginskyy, A. (2017). An open source GIS-

based decision support system for watershed evaluation of best management practices.

Journal of the American Water Resources Association, 53(3), 521-531.

https://doi.org/10.1111/1752-1688.12521

Sidlar, C. L., & Rinner, C. (2007). Analyzing the usability of an argumentation map as a

participatory spatial decision support tool. Urisa, 19(1), 47-55.

https://doi.org/10.1097/01.smj.0000242748.61025.e3

Sieber, R. (2006). Public participation geographic information systems: A literature review and

framework. Annals of the Association of American Geographers, 96(3), 491-507.

https://doi.org/10.1111/j.1467-8306.2006.00702.x

166

Silva, H. De. (2008). Using ICT to reduce transaction costs in agriculture through better

communication: A case study from Sri Lanka. LIRNEasia, Colombo, Sri Lanka, Available

at: http://www.lirneasia.net/wp-content/up- loads/2008/11/transactioncosts.pdf

Simão, A., Densham, P. J., & (Muki) Haklay, M. (2009). Web-based GIS for collaborative

planning and public participation: An application to the strategic planning of wind farm

sites. Journal of Environmental Management, 90(6), 2027-2040.

https://doi.org/10.1016/j.jenvman.2007.08.032

Smithers, J., & Furman, M. (2003). Environmental farm planning in Ontario: Exploring

participation and the endurance of change. Land Use Policy, 20(4), 343-356.

https://doi.org/10.1016/S0264-8377(03)00055-3

Srivastava, P., Hamlett, J. M., Robillard, P. D., & Day, R. L. (2002). Watershed optimization of

best management practices using AnnAGNPS and a genetic algorithm. Water Resources

Research, 38(3), 3-1–3-14. https://doi.org/10.1029/2001WR000365

Stubbs, M. (2014). Conservation reserve program (CRP): Status and issues. Conservation

Provisions in the 2014 Farm Bill and Associated Issues.

Sugumaran, R., Meyer, J. C., & Davis, J. (2004). A Web-based environmental decision support

system (WEDSS) for environmental planning and watershed management. Journal of

Geographical Systems, 6(3), 307-322. https://doi.org/10.1007/s10109-004-0137-0

167

Tang M.Y. (2006). Design and implementation of a GIS-enabled online discussion forum for

participatory planning, MScE thesis, Department of Geodesy and Geomatics Engineering

Technical Report No. 244, Fredericton (NB), University of New Brunswick.

Tang, T. M., & Coleman, D. J. (2008). GeoDF Bridging the communication gap in online

participatory planning. Revue Internationale de Géomatique, 18(4), 443-470.

https://doi.org/10.3166/geo.18.443-470

Togazzini, B. (1992). 1og on Interface. Reading, MA: Addison-Wesley Publishing Company,

Inc.

Trujillo-Barrera, A., Pennings, J. M. E., & Hofenk, D. (2016). Understanding producers’ motives

for adopting sustainable practices: The role of expected rewards, risk perception and risk

tolerance. European Review of Agricultural Economics, 43(3), 359-382.

https://doi.org/10.1093/erae/jbv038

Turpin, N., Bontems, P., Rotillon, G., Bärlund, I., Kaljonen, M., Tattari, S., … Zahm, F. (2005).

AgriBMPWater: Systems approach to environmentally acceptable farming. Environmental

Modelling and Software, 20(2), 187-196. https://doi.org/10.1016/j.envsoft.2003.09.004

Tyllinen, M. et al. (2016). We Need Numbers!: Heuristic Evaluation During Demonstrations

(HED) for Measuring Usability in IT System Procurement’, in Proceedings of the 2016 CHI

Conference on Human Factors in Computing Systems (pp. 4129–4141). doi:

10.1145/2858036.2858570.

168

Udeigwe, T. K., Eze, P. N., Teboh, J. M., & Stietiya, M. H. (2011). Application, chemistry, and

environmental implications of contaminant-immobilization amendments on agricultural soil

and water quality. Environment International, 37(1), 258-267.

https://doi.org/10.1016/j.envint.2010.08.008

Van Haaren, R., & Fthenakis, V. (2011). GIS-based wind farm site selection using spatial multi-

criteria analysis (SMCA): Evaluating the case for New York State. Renewable and

Sustainable Energy Reviews, 15(7), 3332-3340. https://doi.org/10.1016/j.rser.2011.04.010

Vanclay, F. (1992). The social context of farmers’ adoption of environmentally sound farming

practices. In: Lawrence, G., Vanclay, F., Furze, B. (Eds.) Agriculture, Environment and

Society. Melbourne: Macmillan, pp. 94–121.

Veith, T. L., Wolfe, M. L., & Heatwole, C. D. (2003). Optimization procedure for cost effective

BMP placement at a watershed scale. Journal of the American Water Resources

Association, 39(6), 1331-1343. https://doi.org/10.1111/j.1752-1688.2003.tb04421.x

Veith, T. L., Wolfe, M. L., & Heatwole, C. D. (2004). Cost-effective bmp placement:

optimization versus targeting. Transactions of the ASAE, 47(5), 1585-1594.

https://doi.org/10.13031/2013.17636

Veliz, M., Brock, H. and Neary, J. 2006. Ausable Bayfield Conservation Authority watershed

report card 2007. Ausable Bayfield Conservation Authority. Exeter, Ontario. 104 pp.

169

Weiss, P. T., Erickson, A. J., & Gulliver, J. S. (2007). Cost and Pollutant Removal of Storm-

Water Treatment Practices. Journal of Water Resources Planning and Management, 133(3),

218-229. https://doi.org/10.1061/ (asce)0733-9496(2007)133:3(218)

Werts, J. D., Mikhailova, E. A., Post, C. J., & Sharp, J. L. (2011). An integrated WebGIS

framework for volunteered geographic information and social media in soil and water

conservation. Environmental Management, 13(3), 751-767. https://doi.org/10.1007/s00267-

012-9818-5

Westerman, D. (2008). How do people really seek information about others?: Information

seeking across Internet and traditional communication channels. Journal of Computer-

Mediated Communication. https://doi.org/10.1111/j.1083-6101.2008.00418.x

Xie, H., Chen, L., & Shen, Z. (2015). Assessment of agricultural best management practices

using models: Current issues and future perspectives. Water, 7(3), 1088-1108.

https://doi.org/10.3390/w7031088

Yang, W., & Weersink, A. (2004). Cost-effective targeting of riparian buffers. Canadian Journal

of Agricultural Economics, 52(1), 17-34. https://doi.org/10.1111/j.1744-

7976.2004.tb00092.x

Yang, W., Khanna, M., Farnsworth, R., & Önal, H. (2003). Integrating economic, environmental

and GIS modeling to target cost effective land retirement in multiple watersheds. Ecological

Economics, 46(2), 249-267. https://doi.org/10.1016/S0921-8009(03)00141-1

170

Yang, W., Liu, Y., Ou, C., & Gabor, S. (2016). Examining water quality effects of riparian

wetland loss and restoration scenarios in a southern ontario watershed. Journal of

Environmental Management, 174, 26-34. https://doi.org/10.1016/j.jenvman.2016.03.001

Yang, W., Liu, Y.B., Simmons, E.S., Oginskyy, A., McKague, K., 2013. SWAT modelling of

agricultural BMPs and analysis of BMP cost effectiveness in the Gully Creek watershed.

Final Research Report Submitted to Ausable Bayfield, Conservation Authority and Ontario

Ministry of Agriculture and Food

Yang, W., Sheng, C., & Voroney, P. (2005). Spatial targeting of conservation tillage to improve

water quality and carbon retention benefits. Canadian Journal of Agricultural Economics,

53(4), 477-500. https://doi.org/10.1111/j.1744-7976.2005.00031.x

Yiridoe, E. K., Vyn, T. J., Weersink, A., Hooker, D. C., & Swanton, C. (2000). Farm-level

profitability analysis of alternative tillage systems on clay soils. Canadian Journal of Plant

Science, 80(1), 65-73. https://doi.org/10.4141/P99-004

Young, R. A., Onstad, C., Bosch, D., & Anderson, W. (1989). AGNPS: A nonpoint-source

pollution model for evaluating agricultural watersheds. Journal of Soil and Water

Conservation, 44, 168-173. https://doi.org/citeulike-article-id:13481122

Zhang, Y., Sugumaran, R., McBroom, M., DeGroote, J., Kauten, R. L., & Barten, P. K. (2011).

Web-Based Spatial Decision Support System and Watershed Management with a Case

Study. International Journal of Geosciences, 2, 195-203.

https://doi.org/10.4236/ijg.2011.23021

171

Zhang, Y., Wang, L., & Duan, Y. (2016). Agricultural information dissemination using ICTs: A

review and analysis of information dissemination models in China. Information Processing

in Agriculture, 3(1), 17-29. https://doi.org/10.1016/j.inpa.2015.11.002

Zhou, J. & Civco, D.L. (1996). Using genetic learning neural networks for spatial decision

making in GIS. Photogrammetric Engineering and Remote Sensing, 62, 1287–1295.

172

APPENDIX A

“What if” and Policy/Management Scenarios

Table A.1 shows an example of a “What if” BMP scenario script. From the script, three

main user tasks are identified. At first, farmer Bob develops a “What if” scenario. The “What if”

BMP scenario is developed in two steps: 1) select a BMP and 2) assign the selected BMP to

fields. After he develops the scenario, he requests the system to evaluate the environmental and

economic effects of the “What if” BMP scenario. Then, to know the differences before and after

the BMP implementation, he compares the evaluation results of the “What if” BMP scenario

with those of a historical baseline scenario. Clarifying these tasks in the “What if” BMP scenario

is helpful to identify design problems and streamline the design practices.

Table A.1: An example of "What if" scenario

Objective: Describe “What if” BMP scenario

Actors: Bob (a farmer)

Description:

To evaluate a “What if” BMP scenario, Bob chooses BMPs and assigns those BMPs onto

three fields - field 5, field 11, and field 16. Once the BMPs are assigned to his fields, Bob

requests the system to evaluate the environmental and economic effects of the developed

scenario. Bob examines and explores different variables of the modelling results. Then he

selects a historical baseline scenario to examine what the differences in environmental

benefits and economic costs before and after he implements those BMPs are. The

differences reflect the costs, benefits and cost-effectiveness of BMPs.

173

User task steps: Develop a “What if” scenario, Examine and explore the environmental

and economic effects of the “What if” scenario, Compare the scenario with a baseline

scenario to obtain the costs, benefits and cost-effectiveness of BMPs.

Table A.2 and A.3 are generated to derive key user tasks involved in the

policy/management design. Key user tasks include 1) select the policy/management type (either

economic or environmental), 2) select fields in which BMPs will be applied, 3) choose a

pollutant type, 4) specify the economic constraint or environmental target, 5) run the scenario

optimization to obtain the BMP policy/management information, 6) examine the BMP

policy/management information (a combination of BMPs and fields and the corresponding

economic and environmental effects), and 7) investigate the trade-offs between the BMP costs

and the corresponding environmental benefits.

Table A.2: An example of an economic policy/management scenario

Objective: Design economic policy/management for BMP implementation

Actors: Bob (a conservation manager)

Description:

Bob is a conservation manager in a local conservation agency. His current task is to use

the system to design policy/management to implement BMPs to maximize total

phosphorus (TP) reductions under a financial budget constraint. Because the

policy/management design involves a budget constraint, Bob selects the economic

policy/management type. Then, he selects fields to indicate which fields the

policy/management will be applied to. Because the policy/management aims to

maximize TP reductions, he also selects TP as the environmental objective of the

174

policy/management. Based on the selected policy/management type, pollution type, and

fields, the system suggests a range of economic costs for BMP combinations in those

selected fields. Based on the range, Bob specifies a range of budget (economic

constraint) available for implementing BMPs. After he submits the economic constraint,

the system generates a BMP policy/management scenario which includes the BMP

combinations in the selected fields that can maximize environmental benefits (i.e. TP

reductions) under the budget constraints. Also, the system shows the trade-offs between

the economic costs and TP reductions within the specified budget range.

User task steps: Specify the BMP policy/management type; Select fields; Specify the

economic constraint; Run the scenario optimization; Examine the BMP

policy/management information; Examine the trade-offs between the economic costs

and environmental benefits.

Table A.3: An example of an environmental policy/management scenario

Objective: Design environmental policy/management for BMP implementation

Actors: Bob (a conservation manager)

Description:

Bob is a conservation manager in a local conservation agency. His current task is to use

the system to design policy/management to implement BMPs to reduce the TP loadings.

Because the policy/management design involves an environmental target, Bob selects

the environmental policy/management type. Then, he selects fields to indicate which

fields the policy/management will be applied to. Because the policy/management aims

to reduce the TP loadings, he further selects TP as the pollutant type. Based on the

175

selected policy/management type, pollution types, and fields, the system suggests a

number range to indicate the possible TP abatement from BMP combinations in the

fields. Based on the range, Bob specifies a range of TP abatement the

policy/management aims to achieve. After he submits the TP abatement targets, the

system generates a BMP policy/management result to inform him the BMP

combinations in the selected fields that can meet the environmental target and the

corresponding economic costs. Also, the system shows the trade-offs between the

economic costs and TP reductions of BMPs within the specified TP abatement range.

User task steps: Specify the BMP policy/management type; Select fields; Specify

environmental target; Run the scenario optimization service; Examine the BMP

policy/management information; Examine the trade-offs between the economic costs

and environmental benefits.

176

APPENDIX B

Database Table Design for the WebGIS-based Decision Support System for

Facilitating Agricultural BMP Adoption

Table D.1: Structure of the public annotation table

SoilCondition CropType Lat Lon Map Landmanagment Structural

Table D.2: Structure of the news, policy and BMP technical knowledge table

Title Source Content

Table D.3: Structure of the user table

Name Role Password Email Farm ID

Table D.4: Structure of the scenario table

ScenarioName Description UserName ScnearioID CreationDate BMPConfig

Table D.5: Structure of the BMP configuration table

FieldID ConservationTill CoverCrop NutrientManagement WASCoBs

Table D.6: Structure of the modelling result table

FieldID year Flow Sediment Nitrogen Phosphorous Cost Revenue NetReturn

177

Table D.7: Structure of the optimization result table

FieldID BMP NetReturn WaterChange TNChange TPChange SedimentChange

Table D.8: Structure of the discussion topic table

Scenario Date Title Content UserName TopicID

Table D.9: Structure of the discussion thread table

UserName Reply Date ScenarioID TopicID

Table D.10: Structure of the network table

Source Target Weight Role

Table D.11: Structure of the table for “Policy/Management”

BMP_comb Sed_change Water_change TP_change TN_change

NetReturn_change Revenue_change Cost_change

178

APPENDIX C

HttpRequest Sent to Server Using JavaScript

Table B.1: Request to submit an annotation

$.ajax({ url: '/public/writeannotation, type: post, contentType: 'application/json; charset=utf-8', dataType: 'json', data: Annotation, });

Table B.2: Request to search for information

$.ajax({ url: '/public/publicinformationmanagement, type: post, contentType: 'application/json; charset=utf-8', dataType: 'json', data: Keyword, });

Table B.3: Request for user authorization

$.ajax({ url: '/userauthorization', type: 'post', contentType: 'application/json; charset=utf-8', dataType: 'json', data: User, })

Table B.4: Request for scenario creation

$.ajax({ url: '/private/scenariocreation, type: post, contentType: 'application/json; charset=utf-8', dataType: 'json', data: Scenario,

179

});

Table B.5: Request to maintain the BMP assignment relationship

$.ajax({ url: '/private/scenariodevelopment, type: post, contentType: 'application/json; charset=utf-8', dataType: 'json', data: BMPConfiguration, });

Table B.6: Request to run the scenario evaluation

$.ajax({ url: '/private/scenarioevaluation, type: post, contentType: 'application/json; charset=utf-8', dataType: 'json', data: BMPConfiguration, });

Table B.7: Request to update the scenario evaluation chart

$.ajax({ url: '/private/drawevaluationchart, type: post, contentType: 'application/json; charset=utf-8', dataType: 'json', data: ElementType, });

Table B.8: Request to compare two scenarios

$.ajax({ url: '/private/scenariocomparison, type: post, contentType: 'application/json; charset=utf-8', dataType: 'json', data: BaseScenarioName, });

180

Table B.9: Request for scenario optimization

$.ajax({ url: '/private/scenariooptimization, type: post, contentType: 'application/json; charset=utf-8', dataType: 'json', data:OptimizationParameters, });

Table B.10: Request to update the scenario optimization chart

$.ajax({ url: '/private/drawoptimizationchart, type: post, contentType: 'application/json; charset=utf-8', dataType: 'json', data:OptimizationCurveNumble, });

Table B.11: Request to submit the discussion topic

$.ajax({ url: '/private/discussiontopiccreation, type: post, contentType: 'application/json; charset=utf-8', dataType: 'json', data:DiscussionTopic, });

Table B.12: Request to submit the discussion reply

$.ajax({ url: '/private/discussionthreadmanagement, type: post, contentType: 'application/json; charset=utf-8', dataType: 'json', data:DiscussionReply, });

Table B.13: Request for report generation

$.ajax({ url: '/private/reportgeneration, type: post,

181

contentType: 'application/json; charset=utf-8', dataType: 'json', data:ScenarioResults, });

Table B.14: Request for user registration

$.ajax({ url: '/private/userregistration, type: post, contentType: 'application/json; charset=utf-8', dataType: 'json', data:User, });

182

APPENDIX D

Web Services for the WebGIS-based Decision Support System for Facilitating

Agricultural BMP Adoption

Table C.1: WritePublicAnnotation web service

func WritePublicAnnotation(w http.ResponseWriter, r *http.Request, p httprouter.Params) { ConnectToDatabase(annotationDB) WriteAnnotationToDatabse(p.annotation) }

Table C.2: ReadPublicAnnotation web service

func ReadPublicAnnotation(w http.ResponseWriter, r *http.Request, p httprouter.Params) { ConnectToDatabase(annotationDB) ReadAnnotationsFromDatabse(p.mapName) SendAnnotationsToClient(annotationList) }

Table C.3: PublicInformationManagement web service

func PublicInformationManagement(w http.ResponseWriter, r *http.Request, p httprouter.Params) { ConnectToDatabase(publicinformationDB) SearchInformation(p.keyword) SendMessageToClient(searchresult) }

Table C.4: UserAuthorization web service

func UserAuthorization(w http.ResponseWriter, r *http.Request, p httprouter.Params) { ConnectToDatabase(userDB) FindByName(p.name) If name not exist: error message If name exist: If password not correct: error message If password correct: Role(p.name) If user is farmer/manager: direct the user to BMP planning subsystem If user is administrator: direct the user to administration subsystem }

183

Table C.5: ScenarioCreation web service

func ScenarioCreation(w http.ResponseWriter, r *http.Request, p httprouter.Params) { ConnectToDatabase(scenarioDB) WriteScenarioInformationToDatabase(scenarioDB, p.scnearioInfo) RenderTemplate(scenariodevelopmentHTML) }

Table C.6: ScenarioDevelopment web service

func ScenarioDevelopment(w http.ResponseWriter, r *http.Request, p httprouter.Params) { CreateDatabase(BMPConfigurationDB) WriteBMPConfigurationToDatabase(BMPConfigurationDB, p.BMPConfiguration) RenderTemplate(scenarioevaluationHTML) }

Table C.7: ScenarioEvaluation web service

func ScenarioEvaluation(w http.ResponseWriter, r *http.Request, p httprouter.Params) { CreateDatabase(scenarioevaluationresultDB) evaluationResult = Call(integratedmodel, p.BMPConfiguration) WriteEvaluationResultToDatabase(evaluationResult) WriteMapFile(evaluationResult) }

Table C.8: DrawEvaluationChart web service

func DrawEvaluationChart(w http.ResponseWriter, r *http.Request, p httprouter.Params) { elementResult = ReadDatabase(scenarioevaluationresultDB, p.elementType) SendElementResultToClient(evaluationResult) }

Table C.9: ScenarioComparison web service

func DrawEvaluationChart(w http.ResponseWriter, r *http.Request, p httprouter.Params) { basescenarioResult = ReadDatabase(basescenarioDB, p.basescenarioname) comparisonResult = ResultComparison(basescenarioResult, scenarioevaluationResult) WriteMapFile (comparisonResult) }

Table C.10: ScenarioOptimization web service

func ScenarioOptimization(w http.ResponseWriter, r *http.Request, p httprouter.Params) {

184

optimizationResult = Call(programRoutine, p.parameters) CreateDatabase(scenariooptimizationresultDB) WriteEvaluationResultToDatabase(optimizationResult) WriteMapFile (optimizationResult) SendOptimizationResultToClient (optimizationResult) }

Table C.11: DrawOptimizationChart web service

func DrawOptimizationChart(w http.ResponseWriter, r *http.Request, p httprouter.Params) { optimizationchartInfo = ReadDatabase(scenariooptimizationDB, p.optimizationcurvenumber) SendOptimizationResultToClient (optimizationResult) }

Table C.12: DiscussionTopicCreation web service

func DiscussionTopicCreation(w http.ResponseWriter, r *http.Request, p httprouter.Params) { ConnectToDatabase(discussionDB) WriteDiscussionTopicToDatabase(discussionDB, p.topic) SendNotificationEmail(message) }

Table C.13: DiscussionThreadManagement web service

func DiscussionThreadManagement(w http.ResponseWriter, r *http.Request, p httprouter.Params) { ConnectToDatabase(discussionDB) WriteDiscussionReplyToDatabase(discussionDB, p.reply) UpdateCommunicationNetwork (p.replySender, p.replyReciever) SendNotificationEmail(message) }

Table C.14: ReportGenerator web service

func ReportGenerator(w http.ResponseWriter, r *http.Request, p httprouter.Params) { BMPConfiguration = ReadBMPConfigureation(BMPConfigurationDB) If p.reportType == “HTML”: RenderHTMLReport(p.scenarioResult, BMPConfiguration) If p.reportType == “PDF”: text = GenerateTextReport(p.scenarioResult, BMPConfiguration) CallPandoc(text) }

185

Table C.15: UserRegistration web service

func UserRegistration(w http.ResponseWriter, r *http.Request, p httprouter.Params) { ConnectToDatabase(userDB) WriteUserInfo(p.userInfo) SendNotificationEmail(message) }

Table C.16: ScenarioMonitoring web service

func ScenarioManagement(w http.ResponseWriter, r *http.Request, _ httprouter.Params) { tableInfo = ReadTableInfo(discussionDB) SendTableInfoToClient(tableInfo) }

Table C.17: CommunicationManagement web service

func CommunicationManagement (w http.ResponseWriter, r *http.Request, _ httprouter.Params) { networkInfo = ReadTableInfo(discussionDB) SendTableInfoToClient(networkInfo) }

186

APPENDIX E

Evaluation Feedback Scripts

Table E.1: Feedback script examples from evaluation by direct use

Category Measures Feedback scripts System quality Task completeness

& Learnability I finished all the intended tasks. The public subsystem is easy and fun to use. But I found that the BMP planning subsystem a little bit complex. In particular, the procedure of scenario optimization caused some confusion, such as setting the range of constraints.

Task completeness I tested the system performance under a multi-user situation: running the modelling tasks with 20 users at the same time. The system performance was good. No decline in modelling time was noticed compared to the single user situation.

User interface The system interface is well designed. I had no difficulty in navigating the system.

User interface User controls should be added when operations have dependency relationships. For running the optimization model, the “Range” button should be disabled before other parameters were set, i.e. fields, optimization mode, and pollutant type.

User interface The contrast between the button font color and the button background color should be enhanced. I found it hard to read the white font color on the light blue background.

User interface I liked the responsive web page layout, but found the buttons were not arranged properly under small screen resolutions.

User interface The font size may be too small to read. User interface If images are used to indicate entrances of different

modules, it should be clickable to facilitate the system navigation. Tooltips should also be displayed, when the mouse is hovering over the image, to facilitate the system navigation.

User interface Some animations can be removed such as the animation on buttons, because those animations required extra user operations but did not benefit the task process.

User interface The web interface is not fully compliant with the Accessibility for Ontarian with Disabilities (AODA).

187

Improvement is needed at code level (e.g. add attribute to HTML elements).

Information quality

Accuracy The value of the “What if” BMP scenario evaluation results including total phosphorus, total nitrogen, and sediment should be rounded to 3 decimal places.

Clarity I found it difficult to perceive the difference in field-level BMP evaluation results from the map color gradient, which is not in a gradual order. Checking the numerical results was helpful to understand the difference.

Service quality Responsiveness The system loading was fast. Overall the service responded web service requests in a timely manner. The “What if” BMP scenario evaluation took around 1-2 minutes.

Support The user manual was helpful to explain the operation flow to complete the tasks.

Satisfaction Satisfaction I am satisfied with the overall design of the system. It facilitates the use of modelling functions and supports key information tasks for BMP adoption.

Table E.2 Feedback script examples from evaluation during demonstration

Category Measures Feedback scripts System quality Learnability The BMP planning subsystem can be challenging for

farmers to use due to the complexity of the BMP design and evaluation process. Tooltips or other assistance tools should be integrated into the system to guide user operations and improve the learnability of the system.

Task completeness The system provided a set of useful functions to support user tasks in BMP adoption.

User interface The system interface seems easy to navigate. The interface components are well aligned on the web pages.

Information quality

Accuracy The public subsystem allows users to submit public annotations. But mechanisms need to be developed to control or verify the information quality.

Accuracy The BMP planning information generated from integrated economic-hydrologic model and optimization model needs to be accurate.

Clarity The map should display the stream map layers to provide a more complete geographic context.

Interactivity The system offers different ways to present information using maps, tables and charts.

188

Interactions with maps, tables and charts seem straightforward.

Interactivity Explanations to the modelling and optimization results should be added in the system to facilitate users’ understanding and interaction with the results.

Service quality Responsiveness The system modelling takes approximately 1-2 minutes to generate the evaluation result. Given the model simulates ten years of daily data, the running time is acceptable.

Support A well-documented user manual needs to be integrated into the system to explain and guide user tasks by steps. Video tutorial would also be helpful to support the use of the system

Satisfaction Satisfaction The system has a good interface design and offers good functions.

Satisfaction The BMP planning subsystem provides essential and easy-to-access information for supporting decision making on BMP adoptions.

Satisfaction The automated reporting process is very helpful in BMP planning.

Satisfaction The embedded discussion forum and email functions provide a convenient way for farmers to communicate with conservation managers.

Satisfaction The information center in the public subsystem is useful to provide farmers with locale specific information.

Satisfaction The system can be improved by implementing an online community forum for farmers to conduct general discussions and share experiences on BMP adoption.

Satisfaction The system can be improved by incorporating the historical BMP adoption results.

Satisfaction The system can be improved by migrating the public subsystem to the mobile platform.

Impact Awareness The information center in the public subsystem is useful for providing farmers with local specific BMP information.

Awareness The public subsystem, specifically the public annotation sharing site, can be used as a tool to increase farmers’ environmental awareness.

Knowledge The information center in the public subsystem can be used to improve farmers’ knowledge about BMPs.

Knowledge The BMP planning subsystem provides an effective tool to improve the understanding on the economic and environmental effects of BMPs.

189

Intention to use Intention to use The public subsystem has the potential to be used by government staff to collect and monitor information.

Intention to use Future efforts can be planned to train and engage farmers to use this new technology.

Intention to use The system can be used when visiting farmer to discuss BMP adoptions