A GIS-BASED GREEN INFRASTRUCTURE SUITABILTY ANALYSIS …

A GIS-BASED GREEN INFRASTRUCTURE SUITABILTY ANALYSIS FOR STORMWATER MANAGEMENT IN GAINESVILLE, FLORIDA

By

YUXIAO LI

A THESIS PRESENTED TO THE GRADUATE SCHOOL

OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF URBAN OF REGIONAL PLANNING

UNIVERSITY OF FLORIDA

2015

© 2015 Yuxiao Li

To my family and friends

4

ACKNOWLEDGMENTS

First, I want to extend my sincere gratitude to my family and all my friends. Most

significantly, my parents gave me great encouragement to reach the goal to study

abroad. It would not be possible for me to come to University of Florida to finish this

master’s degree without their spiritual and economic support. Also, I would like to thank

my friends who provided pleasures and help during the time studying in University of

Florida.

Second, I am deeply indebted to my thesis committee. It would be impossible for

me to finish this thesis without their help. Dr. Frank, chair of my committee, helped me

go through the hardest part of initiating this thesis topic, and provided the right direction

guidance during my study with patience. Dr. Alakshendra, one of the committee

members, offered valuable suggestions to better accomplish the thesis. Professor

Latimer, one of the committee members, helped me solve the GIS technical problems

during the study.

Third, I would like to thank “Edit24-7”. Its effective and accurate work helped me

solving grammar problems and correcting typographical errors which occurred during

the writing process.

Last but not the least, my gratitude also extends to the Department of Urban and

Regional Planning in University of Florida. You gave me a precious chance to study

abroad to expand my horizon, and granted me unforgettable, joyful and valuable

experience here.

5

TABLE OF CONTENTS page

LIST OF TABLES ............................................................................................................ 7

LIST OF FIGURES .......................................................................................................... 8

ABSTRACT ..................................................................................................................... 9

CHAPTER

1 INTRODUCTION .................................................................................................... 10

Overview ................................................................................................................. 10

Study Area .............................................................................................................. 10

Objective ................................................................................................................. 11

Organization ........................................................................................................... 11

2 LITERATURE REVIEW .......................................................................................... 13

Green Infrastructure ................................................................................................ 13

Benefits of Green Infrastructure .............................................................................. 14

Stormwater Management ........................................................................................ 16

Geographic Information System ............................................................................. 17

Land Use Suitability Analysis .................................................................................. 17

Multi-Criteria Decision-Making (MCDM) Method ..................................................... 18

Analytical Hierarchy Process .................................................................................. 19

Cases...................................................................................................................... 19

Green Infrastructure Planning for Improved Stormwater Management in Central New York, 2012 ............................................................................... 19

Walworth Run Green Infrastructure Feasibility Study ....................................... 21

GIS Suitability Model: Berlin, Maryland ............................................................ 24

Replicable GIS Suitability Model for Stormwater Management and the Urban Heat Island Effect in Dallas, Texas ..................................................... 25

Green Infrastructure Feasibility Scan for Bridgeport and New Haven, Connecticut ................................................................................................... 27

A GIS Suitability Analysis of the Potential for Rooftop Agriculture in New York City........................................................................................................ 28

3 METHODOLOGY.................................................................................................... 29

Definition of Study Area .......................................................................................... 29

Definition of Green Infrastructure ............................................................................ 29

Criteria Selection and Ranking ............................................................................... 29

Analytic Hierarchy Process ..................................................................................... 34

4 PROCESS .............................................................................................................. 37

6

Data Collection and Ranking .................................................................................. 37

Data Processing ..................................................................................................... 38

Criteria Weighing .................................................................................................... 39

Weighted Sum ........................................................................................................ 40

5 RESULTS ............................................................................................................... 45

6 DISCUSSION ......................................................................................................... 53

7 CONCLUSION ........................................................................................................ 56

APPENDIX

A DATA PROCESS .................................................................................................... 57

B RECLASSIFICATION TABLES ............................................................................... 59

LIST OF REFERENCES ............................................................................................... 70

BIOGRAPHICAL SKETCH ............................................................................................ 73

7

LIST OF TABLES

Table page 2-1 Green Infrastructure Benefit ............................................................................... 15

2-2 Development Process of Suitability Analysis ...................................................... 18

2-3 Suitability Ranking Matrix for Rooftop Agriculture in New York City ................... 28

3-1 Green Infrastructure Practice Maximum Slope ................................................... 31

4-1 Suitability Criteria and Ranks.............................................................................. 41

4-2 Pairwise Comparison Matrix ............................................................................... 42

4-3 Normalized pairwise comparison matrix ............................................................. 43

4-4 Criterion Weights ................................................................................................ 44

4-5 Consistency Test Result ..................................................................................... 44

B-1 Soil Type Reclassification ................................................................................... 59

B-2 Slope Reclassification ........................................................................................ 63

B-3 Imperviousness Reclassification ......................................................................... 64

B-4 Land Ownership Reclassification ........................................................................ 68

B-5 Land Use Type Reclassification ......................................................................... 69

8

LIST OF FIGURES

Figure page 1-1 Flowchart of the Whole Research ....................................................................... 12

5-1 Reclassified Soil Map ......................................................................................... 47

5-2 Reclassified Slope Map ...................................................................................... 48

5-3 Reclassified Land Ownership Map ..................................................................... 49

5-4 Reclassified Land Use Map ................................................................................ 50

5-5 Reclassified Imperviousness Map ...................................................................... 51

5-6 Green Infrastructure Suitability Map in Gainesville ............................................. 52

9

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the

Requirements for the Degree of Master of Urban and Regional Planning

A GIS-BASED GREEN INFRASTRUCTURE SUITABILTY ANALYSIS FOR STORMWATER MANAGEMENT IN GAINESVILLE, FLORIDA

By

Yuxiao Li

December 2015

Chair: Kathryn Frank Major: Urban and Regional Planning

Stormwater runoff is one of the biggest challenges to water pollution control for

the reason that this pollution source causes water quality decreases in the U.S. Interest

is now growing in green infrastructure which refers to stormwater management system

mimicking nature by soaking up and storing water. It involves innovative strategies for

addressing the stormwater management problem. Decision-making processes about

green infrastructure implementation can be guided by a green infrastructure suitability

analysis. ArcGIS tools can provide the necessary planning platform for visualization,

modeling, analysis, and collaboration. Analytical hierarchy process provides a strong

evidence for the weight of each criterion in decision making process. So this research is

aiming at developing an appropriate methodology by adapting analytical hierarchy

process for green infrastructure suitability analysis in Gainesville, and creating a

visualized green infrastructure suitability map, integrated with ArcGIS technology, to

highlight areas most suitable for green infrastructures to facilitate green infrastructure

implementation for stormwater management.

10

CHAPTER 1 INTRODUCTION

Overview

Stormwater runoff is one of the biggest challenges to water pollution control

because this pollution source is the main cause of water quality decreases in the U.S.

Interest is now growing in the small scaled green infrastructure which refers to

stormwater manage system to play a same role as nature by soaking up and storing

water. Green infrastructure involves innovative strategies for addressing the stormwater

management problem. Decision-making processes could be guided by a green

infrastructure suitability analysis, which is a tool for finding the most suitable areas for

implementing green infrastructures on the basis of selected criteria. This research on

green infrastructure suitability, in order to solve stormwater management problems, is

integrated with ArcGIS technology, combines two technical requirements (soil type and

percentage slope) and three non-technical requirements (imperviousness, land use type

and land ownership) and adapts analytical hierarchy process which is seldom used in

this field to decide weights of each criterion in order to map green infrastructure

suitability in Gainesville. Figure 1-1 shows the flowchart of the whole research.

Study Area

The study area is the city of Gainesville, the seat of Alachua County, Florida.

Gainesville is the largest city in the county, with an area of 48.2 square miles (City-data,

2015). Florida’s annual rainfall is 53.58 inches, which is 14.4 inches above the national

average (weatherDB, n.d.)

11

Objective

The objective of this research is to develop an appropriate methodology for green

infrastructure suitability analysis in Gainesville, and to create a visualized green

infrastructure suitability map, integrated with ArcGIS technology, to highlight areas most

suitable for green infrastructures to facilitate green infrastructure implementation for

stormwater management.

Organization

The first chapter outlines the purposes of the study. The second reviews the

literature on the topic. The third explains the methodology used. The fourth details the

entire research process. The fifth shows the maps and the results of the research. The

sixth discusses the research and its results. The seventh and final chapter draws the

conclusions of the work.

12

Figure 1-1. Flowchart of the Whole Research

13

CHAPTER 2 LITERATURE REVIEW

Green Infrastructure

There has recently been growing discussion of green infrastructure, especially in

land conservation and development (Benedict & McMahon, 2002). Green infrastructure

is a relatively new term in the urban planning dictionary (Rouse & Bunster-Ossa, 2013).

It has been defined differently, with fine distinctions, by different authors and

organizations. Benedict and McMahon (2012) defined it as “a strategically planned and

managed network of wilderness, parks, greenways, conservation easements, and

working lands with conservation value that supports native species, maintains natural

ecological processes, sustains air and water resources, and contributes to the health

and quality of life for America’s communities and people”. A comprehensive definition

that includes most of these aspects comes from the Conservation Fund, which

describes a green infrastructure as a network instead of a single unit: “Green

infrastructure is our nation’s natural life support system—an interconnected network of

waterways, wetlands, woodlands, wildlife habitats, and other natural areas; greenways,

parks and other conservation lands; working farms, ranches and forests; and wilderness

and other open spaces that support native species, maintain natural ecological

processes, sustain air and water resources and contribute to the health and quality of

life for America’s communities and people” (Benedict & McMahon, 2002). Mell (2010)

examined four definitions of green infrastructure and stated in summary that “access,

spatial variance, multi-functionality, natural and human benefits, biodiversity,

sustainability and connectivity” are the elements that constitute green infrastructure. The

definitions he looked at included almost everything from the large scale, such as

14

vegetation hubs and corridors, to the small scale, such as rain gardens and permeable

pavements. The Environmental Protection Agency (EPA) defines the water quality and

stormwater runoff aspects of green infrastructure. It also classifies green infrastructures

by scale. At the city or county level, a green infrastructure is defined as a patchwork of

natural areas that provides better habitats, air, and water quality, and reduces flood

hazard (EPA, 2014). At the site or neighborhood level, it involves stormwater

management, because a green infrastructure can serve almost the same function as

nature in soaking up and storing water (EPA, 2014). Because past researchers have

used definitions related to their specific research focuses (Mell, 2010) and, in this

research I will use the small scale green infrastructure definition from EPA to further the

study.

Benefits of Green Infrastructure

Green infrastructure benefits human beings and the surrounding areas in several

ways. Kloss (2008) summarized the benefits of green infrastructure for stormwater

management: it can significantly improve water quality by reducing runoff and pollutants

and thus lowering the total contaminants in the receiving water body. Green

infrastructure also mitigates the urban heat island effect, improves air quality, saves

energy, improves the climate, and beautifies communities (Kloss, 2008). A report from

the EPA concluded on the basis of 17 low-impact development (LID) studies that LID

techniques can lower stormwater management costs and enhance environmental

conditions (EPA, 2007). Detwiler (2012) summarized the overall benefits of green

infrastructure in “Growing Green: How Green Infrastructure Can Improve Community

Livability and Public Health” in order to provide resources to help the public and

decision makers understand them. Green infrastructure reduces pollutant runoff into

15

streams and lowers the chance of illnesses being spread through drinking water. A

comprehensive and concise summary was published by EPA (2010) in “Green

Infrastructure Case Studies: Municipal Policies for Managing Stormwater with Green

Infrastructure,” which found three kinds of benefit: environmental, economic, and social.

The summary is showed in Table 2-1.

Table 2-1. Green Infrastructure Benefit (EPA, 2010)

Environmental Economic Social

Benefit

Increase carbon sequestration

Improve air quality

Additional recreational space

Efficient land use Improve human

health Flood protection Drinking water

source protection

Replenish groundwater

Improve watershed health

Protect or restore wildlife habitat

Reduce sewer overflow events

Restore impaired waters

Meet regulatory requirements for receiving waters

Reduce hard infrastructure construction costs

Maintain aging infrastructure

Increase land values

Encourage economic development

Reduce energy consumption and costs

Increase life cycle cost savings

Establish urban greenways Provide pedestrian and

bicycle access Create attractive

streetscapes and rooftops that enhance livability and urban green space

Educate the public about their role in stormwater management

Urban heat island mitigation

16

Stormwater Management

Stormwater is a widely used term not only in scientific documents but in

regulatory reports. It is one of the starting points of this thesis. Stormwater has been

treated as the primary contributor to water quality impairment, but federal regulations

have been passed in the last 20 years to address this problem (National Research

Council, 2009). Stormwater runoff is defined by the National Research Council (2009)

as “the water associated with a rain or snow storm that can be measured in a

downstream river, stream, ditch, gutter, or pipe shortly after the precipitation has

reached the ground”.

Between a third and a half of the transformation of the earth’s land surface is due

not simply to the growth of the human population but to human activities (Vitousek,

Mooney, Lubchenco, & Melillo, 1997). Land coverage and drainage systems not only

affect stormwater flow, they influence the quantity of baseflow between storms

(Randolph, 2004). The urbanization process has led to increments in the peak

discharge by limiting water infiltration into the ground and thus increasing the flow rate

during the runoff accumulation period (Randolph, 2004). Stormwater is creating

management problems for cities in U.S. and transforming urban water systems, which

can be regarded as a reason for putting an end to pollution (Kloss, 2008). According to

the National Water-Quality Assessment (NAWQA) program from the U.S. Geological

Survey (USGS), the total phosphorus concentration in 70 percent of urban steams

exceeds the EPA’s goals for nuisance plant growth control; higher concentrations of

insecticides occur in urban steams than in agricultural areas; and fecal coliform bacteria

in urban streams exceed the standard limits for safe water recreation (USGS, 2008). In

recent years, because of the objective expansion and improvement of planning

17

methods, stormwater management has been changing tremendously (Randolph, 2004).

Of particular note is low-impact stormwater management, which involves controlling and

treating onsite or decentralized runoff more effectively, advocating or encouraging LID

designs and integrated stormwater controls, and restoring natural channels through

bioengineering and infiltration (Randolph, 2004).

Geographic Information System

A geographic information system (GIS) has been defined as “a set of tools for the

input, storage and retrieval, manipulation and analysis, and output of spatial data”

(Malczewski, 1999). The National Geographic Society (2011) defines GIS as “a

computer system for capturing, storing, checking, and displaying data related to

positions on Earth’s surface”. GIS can provide strong visual displays of analysis results

at multiple scales, and it is an excellent technology for researchers or planners to gain a

thorough understanding of their problems and to better solve them by incorporating with

data with specific attributes (Berger, 2013).

Land Use Suitability Analysis

Land use suitability analysis is an identification tool for locating the most suitable

land for future use (Collins, Steiner, & Rushman, 2001). Collins, Steiner, and Rushman

(2001) introduced the land-use suitability analysis and discussed its development in

terms of both methodology and technology in the U.S. Table 2-2 shows the

development process of suitability analysis. GIS-based land use suitability analysis,

according to Malczewski (2004), is differentiated into three main approach groups,

which are similar to the last three stages in the table above: computer-assisted overlay

mapping, multi-criteria evaluation methods, and AI (soft computing or geo-computation)

methods.

18

Table 2-2. Development Process of Suitability Analysis (Malczewski, 2004)

Stage 1 Stage 2 Stage 3 Stage 4 Stage 5

Name Hand-drawn map

Literature advancement

Computer- assisted map overlay

Spatial data and multi-criteria evaluation redefinition

Replicating knowledge

Details Overlay hand-drawn map manually.

Draw map on transparent papers with the same scale and control features.

Apply computer technology in land-use suitability analysis.

1. Boolean logic-handling indiscriminant spatial boundary with GIS.

2. Decision makers’ preference incorporation with land-use allocation.

Artificial intelligence integration for model applications and GIS-based land use allocations.

Multi-Criteria Decision-Making (MCDM) Method

The multi-criteria decision method (or multi-criteria decision analysis) is used to

make decisions incorporating the decision makers’ explicit preferences, represented by

goals, constraints, weights, and similar parameters (Al-Shalabi, Mansor, Ahmed, &

Shiriff, 2006). Multi-criteria decision making, which is a part of the methodology of this

thesis, when integrated with GIS, is far more effective than the traditional map-overlay

method for land-use suitability analysis (Carver, 1991). The method has been used this

way by researchers in many different places. GIS-based MCDM is the process of

combining and transforming spatial elements, which are the input data, into final

decisions as output data (Malczewski, 2004). Spatial MCDM, according to Malczewski

(2004), needs to consider two significant elements. The first is the capability for GIS-

data acquisition, storage, recovery, manipulation and analysis. The second is the

capability for combining the data and the decision makers’ preferences, which are to be

unified in dimensional value form to produce decisions (Malczewski, 2004).

19

Analytical Hierarchy Process

One of the main problems in the MCDM process is setting the weights of the

different criteria according to their importance. The AHP method can solve this problem,

according to Banai-Kashani (1989), because it can detect and correct errors in

judgements of factor importance during site suitability analysis. AHP has been used in

many suitability analysis studies. As an MCDM approach, AHP is used widely in

agricultural land-use suitability analysis (Akıncı, Özalp, & Turgut, 2013). Al-Shalabi,

Mansor, Ahmed, and Shiriff (2006) did a housing-site suitability analysis using the AHP

methodology. Duc (2006) did a land-use suitability analysis for coffee using GIS and

AHP in the Ha district. Kumar and Shaikh (2013) used this technique in combination

with GIS to process a site-suitability assessment for Mussoorie municipal area

development. This thesis will adopt AHP as part of its methodology for dealing with the

criterion-weighing problem.

Cases

Several cities have analyzed green infrastructure suitability or feasibility. They

undertook their studies with different goals, used different data, chose criteria from

different aspects, and accomplished the analysis with different methods. This thesis was

inspired by them, especially in the criterion selection part.

Green Infrastructure Planning for Improved Stormwater Management in Central New York, 2012 (Central New York Regional Planning & Development Board, 2012)

New York City carried out its suitability analysis mainly to test the viability of 18

stormwater practices. This study also graphically illustrated some of the factors that

affect the decision to consider or disregard specific stormwater practices in specific

areas. Six geographic factors were selected for the NYC GI-suitability analysis. These

20

were hydrologic soil group, land use, slope, proximity to roads, presence or proximity of

wetlands, and floodplains. The suitability value for each factor ranged from 0 (least

suitable) to 5 (most suitable). It should be noted that although the analyses were made

using the same geographic factors, the ranking method varied according to specific

designs.

The NYC study classified soil into four hydrologic groups in accordance with

USDA Natural Resources Conservation Service data. These were: very permeable

sandy or sandy loam soils, loams or soils with a high percentage of silt, loams with high

percentages of both sand and clay, and nearly impermeable clays or clay loams. This

analysis rated soil as the most important factor by giving it a weight of 0.3 out of 1.

Green infrastructure suitability is affected by the area’s land use character. This

factor was broken into two components: the type of development, and the perception of

the practice by the public in the context of the current land use. For example, certain

practices (e.g., stormwater wetlands, infiltration, trenches, and sandy filters) are

considered less suitable in residential areas because residents are thought to dislike

their appearances or the effects they might have, such as providing breeding areas for

mosquitoes. Residential areas were thus rated low. On the other hand, commercial

areas were rated high for pervious pavements and bio-retentions because of their vast

parking areas. This analysis gave a weight of 0.25 out of 1 to land-use type.

Slope is a core consideration for the majority of kinds of green infrastructure. This

analysis defined slopes of 1–5% and 2–10% as high suitability because some practices

can be implemented correctly and function better on flatter ground. Areas with slopes

exceeding 15% receive scores of 0. Slope was weighted at 0.2 out of 1 in this analysis.

21

Proximity to roads was taken into account because roads provide easier access

for inspecting and maintaining large-scale green infrastructures. Smaller-scale practices

(rain gardens, rain barrels, and stormwater plants) do not have the same requirements

for heavy on-site maintenance equipment, however. So different practices were

analyzed on different scales. This factor was weighted at 0.1 out of 1.

Some green infrastructures should not be implemented in wetlands or hydric soil

areas, so this factor needed to be taken into consideration in the analysis. For practices

relying on infiltration for runoff reduction or water quality improvement, such as pervious

pavement and bio-retention, wetlands have a low suitability. However, they have a high

suitability for practices like stormwater wetlands that rely on the surrounding water. It

was weighted at 0.1 out of 1.

Most green infrastructures are not ideally located in floodplains. These practices

(such as permeable pavement, bio-retention, and swales), will be given suitability

scores in the 1 to 4 range. The presence of floodplain is not very important, however,

since green infrastructure practices are meant to deal with small volumes rather than

the 100-year storms that the floodplain designation is based on. So this factor was

weighted at only 0.05 out of 1 (Central New York Regional Planning & Development

Board, 2012).

Walworth Run Green Infrastructure Feasibility Study (Northeast Ohio Regional Sewer District, 2011)

This study is different from the one above. It was influenced by a local greenway

plan intended to revitalize the Walworth Run stream corridor in the Stockyards

neighborhood, and its emphasis is on the stream that carries the Walworth Run sewer

out to the Cuyahoga River, CSO 080. It was aimed at educating the neighborhood on

22

the benefits of green infrastructure to the surroundings and illustrating how this analysis

helped with the enhancement process. A unique feature of this feasibility study is the

defining of subsheds. The district set numerous goals for reducing combined sewer

overflow (CSO) volume, and the sewershed sub-division process made for more

effective evaluation. A subshed is defined as a catchment aggregation based on natural

divisions. The study area was divided into 42 subsheds ranging in size from 30 to 150

acres. Six criteria were selected and analyzed by subshed. The analysis process

involved assessing each subshed’s characteristics for each criterion, and then ranking

all the criteria, from 1 (low) to 3 (high), according to the condition of each subshed. After

that, different multipliers were applied to each criterion to calculate the final feasibility.

Redevelopment coordination (*5). This factor was included in the analysis

because the district had to complete 42 million dollars in green infrastructure projects

within eight years and needed to incorporate them with its existing projects to stimulate

economic development. Thus 3 points were given to individual projects constructed

within 5 years; 2 points to projects constructed in 5 to 10 years, and 1 point to projects

requiring longer than that. After the scores were calculated for each subshed, a score of

0–4 points was regarded as low feasibility, 4–6 points as medium feasibility, and 7–12

points as high feasibility.

Vacant and landbank properties (*5). Green infrastructures can mitigate the

negative effects of vacant land and improve the quality of life for the residents of the

neighborhoods in Cleveland. Vacant land and landbank property are also thought to

benefit neighborhoods by increasing land value. The larger the vacant area is, the more

flexible it Is for large-scale green infrastructure implementation. Hence, vacant

23

properties that were larger than 2 acres received 3 points. Those between three-

quarters of an acre and 2 acres received 2 points, and smaller areas received 1 point.

After computing the points in each subshed, the district categorized the result as

follows: 0–4 points meant low suitability, 4–9 meant medium suitability, and 10–19

points meant high suitability.

Public lands adjacent to vacant and landbank properties (*2). The district

judged that identifying a partner whose mission was consistent with the district’s own

concentration on clean water would be helpful to green infrastructure implementation.

The potential partners included school properties, parks, and non-profit properties. After

establishing the partner layer by putting all the elements together, the District overlaid

the partner layer with the vacant or landbank property map to determine the feasibility

for these criteria. Subshed-owning partnerships adjacent to the vacant or landbank

property offered a high feasibility. Subshed-possessing partnerships within 500 feet of

the property were medium feasibility. The rest were low feasibility.

Impervious areas (*3). An effective way of reducing CSO volume is to prevent

stormwater from reaching the combined sewer systems by having a large number of

impervious areas connected to the sewer system. By tallying the parking lots and large

building areas in each subshed, the district was able to reclassify the impervious area

as follows: subsheds with less than 5% impervious area were low suitability; subsheds

with 5–10% impervious area were medium suitability; and those with more were high

suitability.

Minorities and poverty (*1). Green infrastructures can provide an overall

socioeconomic improvement to a community. In this case, the 33% minority-population-

24

rate map and the 13% poverty-rate map were overlaid. Subsheds with both of these

characteristics were viewed as high feasibility. Subsheds with just one of the two

categories were deemed medium feasibility. Those with neither were regarded as low

feasibility.

Soils (*1). The urbanization process in Cleveland caused soil displacement,

which affected the sites’ capacity for water infiltration. Because the urban condition was

complex and the site specifics were unknown, the district classified soil into only two

types: subsheds where historic soil maps showed a sandy condition were high

feasibility, and subsheds where the maps showed potential soil restrictions were low

feasibility.

GIS Suitability Model: Berlin, Maryland (Marney, 2012)

This analysis was done in a thesis to determine green infrastructure suitability on

the basis of GIS data to deal with stormwater in Berlin. My research will use the same

definition of green infrastructure, which is a system mimicking natural processes. This

research also sorted green infrastructures into rain gardens, filter strips, bio-swale,

permeable pavement, and bio-retention ponds. It then analyzed the city’s

implementation requirements and selected slope, soil type, proximity to structures, land

ownership, depth to ground water, and land use. The criteria selection process was

based mainly on New York City’s green infrastructure suitability analysis. The rank

values were 1 (low) through 3 (high). Slopes of 2–4.9% were ranked high, slopes of 5–

8% medium, and slopes of 0–1.9% low. The soil type factor was ranked the same way

as in New York. Land use was categorized as commercial, industrial, or residential,

which were deemed to have high, medium, and low suitability respectively. These three

factors were also taken from New York’s analysis. Proximity to structures was included

25

in order to avoid damage to structure foundations. A distance of more than 15 feet to

the nearest structure meant high suitability, with a score of 3. Distances of 10 to 14.9

feet and of 9.9 feet or less were regarded as medium and low suitability, respectively.

Land ownership was considered, according to Marney (2012), because it influences the

implementation process. Private landowners are likely to create obstacles to green

infrastructure implementation. Private land was thus given a low-suitability score of 1.

Public land, on the other hand, was given a 3. Depth to ground water, a unique criterion

here, was considered according to the requirements of the particular tools (Marney,

2012). A depth of more than 5 feet meant high suitability and a score of 3. Depths of 2

to 4.9 feet and of 1.9 feet or less get scores of 2 and 0 respectively.

Replicable GIS Suitability Model for Stormwater Management and the Urban Heat Island Effect in Dallas, Texas (Buchholz, 2013)

The aim of this study, also a graduate student thesis, was to promote green

infrastructure and LID implementations for the stormwater management and urban heat

island problems by adopting a replicable GIS suitability model. The study selected five

criteria for determining suitability. Each criterion was given a score from 1 (low

suitability) to 3 (high suitability).

The minority-and-poverty status criterion was taken from the Walworth Run study

due to the similar demographic situation: Dallas has a large middle-class Latino

population. Areas with more than 40% poverty and more than 64.9% minority

populations were defined as highly suitable and given a score of 3. Areas with only one

of those characteristics were given a 2, areas with neither were defined as low suitability

and given a 1.

26

Soil was also included in this study. In fact, soil and slope were analyzed

together because the data contained them together. Eighty types of soil were sorted into

three main groups. The ranking method accorded largely with that of the Berlin study

and the EPA’s urban soil suitability evaluation for green infrastructure. Sandy soil and

sandy-loamy soil were given a score of 3, representing high suitability. Loamy soil was

given a 2, and clay a 1, representing low suitability. Soil with a slope of 0–1% or of more

than 8% was given a 1.

Land surface temperature was used to indicate hot spots, areas that would

benefit from green infrastructure. This factor was taken into consideration mainly on the

grounds that green infrastructure can mitigate the urban island heat effect. The hot spot

identification process involved classifying raster values of temperature with a 0.5

standard deviation. Top stand deviation areas were regarded as having high suitability

and the rest as low suitability.

Land cover was selected for its effect on land surface temperature. The more

impervious an area was, the more it would benefit from green infrastructures. Areas

were sorted into three groups on the basis of this criterion. Fifty to 100% impervious

developed land cover indicated high suitability and received a score of 3; 0–49%

impervious developed land cover meant medium suitability and a score of 2; natural and

agricultural land cover meant low suitability and a score of 1.

Tree canopy data was used to better understand the recent environmental

conditions in Dallas. Areas with less tree canopy could benefit more from green

infrastructure. An area with a 0% tree canopy received a score of 3 (high suitability),

with a 1–80% canopy, a 2 (medium), and with an 81–100% canopy, a 1 (low).

27

This analysis used two models. One was used to give each criterion the same

weight (0.2). The second was used to give the minority and poverty and land

temperature criteria more weight (0.25), and the tree canopy criterion less (0.1), with the

rest the same. The second model appeared to be more selective in identifying high-

suitability areas for green infrastructures, and naturally this was adopted for the result of

the analysis.

Green Infrastructure Feasibility Scan for Bridgeport and New Haven, Connecticut (City of New Haven, 2012)

Bridgeport and New Haven conducted a feasibility study in order to evaluate

opportunities for green infrastructure incorporation. The main purpose of the study was

to function as a basis for future planning and design. It differed from the studies

discussed above in that it merely enumerated factors that might influence green

infrastructure site selection without ranking or weighing them. They were still useful in

the criteria selection for the present study.

Publicly owned areas were considered better sites for green infrastructure

implementation because they allowed for reduction not only in the cost but in the

complexity of the whole process, including planning, constructing, and maintenance.

Furthermore, green infrastructures implemented on public land could better serve as

educational tools. Green infrastructures implemented in CSO areas could also play a

better role in source control. Area presence was also considered as a factor. Sites with

better visibility could both display the efforts of the city to improve water quality and

function as evidence of how well they performed and how they benefitted the city. As

with the Walworth Run study, incorporation with existing developments was included

because of the importance of reducing expenditures on the more expensive types of

28

green infrastructure implementation. The final criterion was the difficulty in each area of

making modifications for source control. Areas where little modification was needed to

gain the benefits of green infrastructure were considered highly suitable for green

infrastructure implementation.

A GIS Suitability Analysis of the Potential for Rooftop Agriculture in New York City (Berger, 2013)

This research led me to exclude green roofs, green walls, and similar systems

from my study. It used a completely different method from the other studies. The main

reason is that this kind of design tool is not implemented on the ground; it is adapted on

or in the building. The methodology was basically a building selection process. The first

step was determining the area of interest by selecting for buildings of 10 floors or fewer

occupying more than 10,000 square feet, completed before 1968, and located in

commercial and manufacturing zones. The area of interest was refined as Brooklyn.

The only ranking criteria were the area of each building and the year it was built. Table

2-3 shows how the buildings were ranked. A score of 3 to 6 represents one specific kind

of green roof. 1 and 2 indicate a low potential for green roofs.

Table 2-3. Suitability Ranking Matrix for Rooftop Agriculture in New York City

Suitability ranking

matrix

Prior to 1968

(Intensive)

After 1968

(extensive)

Small scale

0–5000 sq.ft. 2 1

Medium scale

5000–40000 sq.ft. 5 3

Large scale

More than 40,000 sq.ft. 4 6

29

CHAPTER 3 METHODOLOGY

Definition of Study Area

The study area of this thesis will be the city of Gainesville in Alachua County,

Florida. Gainesville is the largest city in the county, with an area of 48.2 square miles

(City-data, 2015). The estimate population in 2014 is 128460. The elevation of

Gainesville is 121 feet. Florida’s annual rainfall is 53.58 inches, which is 14.4 inches

above the national average (WeatherDB, n.d.). Over the last 30 years, Gainesville

averaged 47.33 inches annually, which is 8.16 inches above the national average.

Rainfall was heaviest in the summer, with 19.58 inches.

Definition of Green Infrastructure

Green infrastructure has been defined differently throughout the literature. At the

city or county level, it generally means the patchwork of natural areas that provide

habitat, flood protection, clean air, and clean water. At the neighborhood or site level, it

involves stormwater management systems that mimic nature by soaking up and storing

water (EPA, 2014). In this thesis, a small-scale definition of green infrastructure will be

used. It will include these categories: bio-swales, tree cover, rain gardens, permeable

pavements, bio-retention ponds, and other, similar designs. Green roofs and green

walls will be excluded because the methodology of the green roof suitability analysis

study discussed in the literature review was different from the others.

Criteria Selection and Ranking

Criteria selection is an important part of this research. The criteria selection

principle is mainly based on past suitability analyses. Three main considerations

determined the factors used here: technical requirements, non-technical factors, and

http://rainfall.weatherdb.com/l/63/Gainesville-Florida

30

data availability. Suitability is divided into 3 group: high suitability, medium suitability and

low suitability. This study will use score of 3, 2 and 1 to indicate high suitability, medium

suitability and low suitability respectively.

Among the technical requirements, soil and slope were both selected as in

several past studies. Soil type has been used in suitability analyses in many cities. Soil

is made up of different components, and different soil types have different particle sizes.

Green infrastructure tests should concentrate not only on the soil’s capacity for

stormwater runoff retention and infiltration, but on its support for natural vegetation

(EPA, 2011). This is how I have ranked this criterion.

Sandy soil has the ability to keep its structure and does not create serious

drainage or compaction problems. It is less likely to have high levels of soluble

contaminants, and it responds well to organic additions. Green infrastructures function

well on sandy soils for stormwater infiltration work (EPA, 2011). Sandy soils will thus be

given a high score of 3. Loamy soils are blends of sand, clay, and silt that show each

property evenly. They are the best soils for producing high yields from food crops. Their

drainage and infiltration properties depend on the relative proportions of their

component materials. Drainage works well for maintaining the ideal soil moisture levels

for plant growth in uncompacted loamy soils (EPA, 2011). Loamy soils will be given a

score of 2. Clayey soil’s particle size is small and flat. Because of that, its capacity for

absorbing water is good. Water erosion on clayey soil is often severe, however, and

clayey soils are not typically well drained. Natural phenomena and human activity can

easily compact clayey soil. Aeration and soil amendment are generally needed to

31

establish vegetation on clayey soil (EPA, 2011). Therefore, clayey soil will be given a

low score of 1. Water will receive a score of 0.

Slope is considered not only because it appears in the past studies but because

it affects the functioning of green infrastructures. Table 3-1 contains several green

infrastructure practices and the maximum slope for each as recommended by various

sources (EPA, 2014). Several green infrastructures are limited to slopes of less than

6%, so slopes of 0–6% will be given a high score of 3. Slopes of 6–10% will be given a

2, and those of 10% or more will be given a low score of 1.

Table 3-1. Green Infrastructure Practice Maximum Slope (EPA, 2014)

Green infrastructure practice

Maximum Slope Reference

Bioretention/Vegetated swale/Planter box

6% CWP, 2009: Penn. SW BMP Manual (BMP 6.4.5/BMP 6.4.8)

Dry well 6% CWP, 2009: Penn. SW BMP Manual (BMP 6.4.6)

Grass Channel with check dams (vegetated swale)

5% CWP, 2009: CED Engineering website

6% CWP, 2009: Penn. SW BMP Manual (BMP 6.4.8)

Diversion/Infiltration berm (terracing)

25% CWP, 2009: Penn. SW BMP Manual (BMP 6.4.10)

Infiltration trench 5% CWP, 2009: Penn. SW BMP Manual (BMP 6.4.4)

Permeable pavement 5% Fassman and Blackbourn, 2010; CWP, 2009; Muench et al. 2011: Penn. SW BMP Manual(BMP 6.4.1)

Vegetated filter strip 6% CWP, 2009 8% Penn. SW BMP Manual

(BMP 6.4.9) 33% Navickis-Brasch, 2011

Non-technical factors are imperviousness, land ownership, and land use. Land

cover (imperviousness) was selected by both Walworth Run and Dallas to analyze

suitability. Data on intact impervious surfaces can make the model more precise

32

(Marney, 2012). Some of the impervious areas are connected directly to the sewer

systems, so that stormwater falling on the impervious areas will flow directly into the

sewers (Northeast Ohio Regional Sewer District, 2011). Thus the more impervious the

land is, the more area is needed to implement green infrastructures. Both of the other

studies selected land cover (imperviousness) as a parameter; however, they used

different methods to rank it. Walworth Run defined imperviousness by subshed, which is

catchment aggregation based on divisions created by land use, neighborhoods, and

roadways (Northeast Ohio Regional Sewer District, 2011). Walworth Run calculated

imperviousness from the percentage of the land in each subshed covered by parking

lots and large buildings (Northeast Ohio Regional Sewer District, 2011). Dallas, on the

other hand, simply based its study on the 2006 National Land Cover (NLCD) dataset,

which gives the imperviousness of the land directly. Dallas categorized land into four

kinds of imperviousness: water; 0–20% impervious land cover or natural/agricultural

land cover; 20–49% impervious land cover; and 50%—100 impervious land cover

(Buchholz, 2013). Taking only parking lots and large buildings into consideration yields

less accurate results than the Nation Land Cover dataset. In this thesis, I will use the

same method as Dallas for categorizing and ranking imperviousness. Green

infrastructures benefit impervious areas more (Buchholz, 2013). Developed land that is

50–100% impervious has high suitability and will get a score of 3; 20–49% impervious

land will get a 2; and natural land cover and 0–20% impervious developed land has low

suitability and will get a score of 1. Water will be scored 0 because of the impossibility of

implementation. The difference in this ranking method is the treatment of 0–20%

33

impervious developed land. These lands are mostly vegetation in the form of lawn

grasses. It is more appropriate to rank them as low in suitability.

Land ownership was selected as a parameter because of social factors in Berlin

and in Bridgeport and New Haven. Berlin considered the divergence between public

and private interests. Land owned by the public can be easily fitted with green

infrastructures; private land, however, can be difficult to adapt depending on the owner’s

preferences (Marney, 2012). Private land can thus create obstacles to the green

infrastructure implementation process. In Bridgeport and New Haven, ownership was

also considered for educational purposes. Green infrastructure projects provide

opportunities to stakeholders to understand green infrastructure and its implementation

issues in the real world (City of New Haven, 2012), and in public areas they allow the

public to see what green infrastructures look like and how they work (City of New

Haven, 2012). Professional staff can also benefit from green infrastructures in public

areas by gaining real world experiences during the design, implementation, and

maintenance processes (City of New Haven, 2012). In addition, a key objective of

Alachua County’s updated comprehensive plan was to improve ways of managing

conflicts between developmental, agricultural, and environmental uses (Alachua County

Growth Management Department, 2011). This indicates that land management is a

problem in Alachua County. For example, private land can create barriers to green

infrastructure implementation. Consequently, community-level, rather than individual-

level implementation of green infrastructure can help local governments achieve their

environmental, sustainability, and adaptation goals (Foster, Lowe, & Winkelman, 2011).

34

Thus public land has high suitability and receives a score of 3; private land has low

suitability and a score of 1.

Land use was chosen as a factor for social reasons. New York City and Berlin

selected it as a criterion because residents often dislike the appearance of green

infrastructures. Also, even though the biological and chemical processes of some green

infrastructures can take up some nutrients, such as phosphorus, from stormwater, the

depression areas that become inundated with plants can cause mosquito breeding

problems (Rector, Duckworth, & Obropta, 2012), which residents dislike. Bio-retention

and pervious pavements can work well in commercial areas because of their

widespread parking lots. In Gainesville in particular, where University of Florida is

located, high suitability should give to school areas, which are classified as institutional

lands, because green infrastructures can be used as educational resources for the

students there. Their educational value can be better realized when they are located

near schools. Another major area of institutional land in Gainesville is the airport.

Implementing green infrastructure on airport land can be a very sustainable method of

stormwater management. Thus residential areas will be given a low score of 1, while

commercial and institutional areas will be given a high score of 3. Other kinds of land

will be given a score of 2.

Analytic Hierarchy Process

The analytic hierarchy process (AHP) is a way of deriving relative scales from

judgment or from data from a standard scale and determining how to perform the

subsequent arithmetic on these scales to avoid unnecessary number crunching. Past

researches about green infrastructure suitability analysis have seldom used this method

to determine weights for criteria. Also, some of the researches claims that criteria

35

weighting is one of their shortages of using empirical way. This research adapts analytic

hierarchy process method which provides a strong evidence to determine weights for

criteria.

For a suitability analysis, it is necessary to give a score to each criterion. For this

purpose, Saaty's nine-point weighing scale was applied in a pairwise comparison

matrix. To develop a pairwise comparison matrix, different criteria are required to create

a ratio matrix. The pairwise comparisons are then taken as input and relative weights

are produced as output. The weights on specific criteria can be calculated after the

formation of the pairwise comparison matrix. The first step is finding the sum value in

each column in the matrix. The second step is normalization, which uses the column

total to divide every element in the matrix. The last step is to compute the average of

each row in the normalized matrix. This average will be the weight of the corresponding

criterion. People do not innately use logic consistently every time (Saaty, 1994), so after

the weight of each criterion has been computed, the consistency ratio (CR) should also

be measured to eliminate bias. A reasonable value for this is a CR of less than 0.1. A

greater value indicates that the weights are inconsistent. In this case the pairwise

comparison matrix should be revised and reconsidered (Kumar & Shaikh, 2012). The

CR can be calculated by the formula

CR = CI/RI (3-1)

in which CI is the consistency index. CI can be computed, under the condition that λ is

greater than or equals to n, by the formula

CI = (λ-n)/(n-1). (3-2)

36

Here, λ is the eigenvalue of the matrix and n is the criterion number in the research. RI

is the random index, which for this research, where the number of criteria is 5, RI equals

1.12 (Saaty, 2000).

37

CHAPTER 4 PROCESS

Data Collection and Ranking

Determining the criteria, which are soil type, slope, imperviousness, land

ownership, and land use, is the first step. The next step is to collect the relevant data.

For soil type, I used the soil data from the Natural Resources Conservation

Service (NRCS). This is collected through the National Cooperative Soil Survey, and it

contains the most detailed soil geographic data. The geodataset type is shapefile and

the geodataset feature is polygon, with attributes on each unit of land. The most useful

information for this research was the map unit attribute, which is named “MUNAME” in

the attribute table. It contains the information on the soil characteristics of each unit.

The slope data could not be gathered directly. However, ArcMap has a tool

called slope that can be used accomplish this step by incorporating digital elevation

model data. Florida’s statewide 5-meter digital elevation model (DEM) was used to

collect data. This is not the only reason I selected this dataset, however. To keep

process consistent, all the raster data and processes were set with a 5-meter cell size.

The DEM dataset just fit the consistency objective.

Imperviousness data comes from “NLCD 2006 Land Cover (2011 Edition,

amended 2014)”. The NLCD (National Land Cover Database) products are created as

part of a cooperative project conducted by the Multi-Resolution Land Characteristics

(MRLC) Consortium. The dataset consists of raster data presented in a remote-sensing

image mode. The attribute table does not show the attribute information directly, but

only the number information. However, the metadata explains what each number

38

means. Detailed information is in the appendix. The ranking process is based on what

the dataset contains and how it represents the information.

Land ownership data was obtained from “ALACHUA COUNTY FLORIDA

PARCEL DATA – 2010.” The purpose of the data is “to serve as base information for

use in GIS systems for a variety of planning and analytical purposes.” It consists of

vector data with attributes on each unit (parcel). The geodataset type is shapefile and

the geodataset feature is polygon. The public land code is listed in the attribute table for

each parcel. I use this information to distinguish public and private land.

Land-use data was acquired from “FDOT DISTRICT 2 – GENERALIZED

FUTURE LAND USE.” This contains information on both future land use and current

land use of each unit, represented by a vector type. The ranking process is based on

what the data can offer to further the research.

The collected data also included information on the study area boundaries. The

study area is Gainesville, in Alachua County, so I used the “CITY LIMITS – DERIVED

FROM FLORIDA PARCEL DATA – 2011” data set, which contains all the city limits in

Florida. This dataset will be used as a limitation tool for the research.

The main reason for the ranking process is given in the methodology section

above. Table 4-1 below summarizes how each criterion was ranked.

Data Processing

The first step of the process was to isolate Gainesville’s city limits in the Florida

city limits dataset. The collected data did not cover Gainesville alone; some of them

covered the whole county or state. Because the study area is Gainesville, the data for

each criterion were masked using the isolated Gainesville city limit layer. To keep the

research consistent, the raster data processing cell size was 5 meters, the processing

39

mask was Gainesville city limits, and the processing extent was also Gainesville city

limits. Another significant measure to ensure the consistency of the process was that all

the raster data that were transformed from vector data were extracted using

Gainesville’s boundaries instead of being clipped before the transformation process.

The data processing steps are shown in detail in the appendix. The maps appear in the

Results section.

Criteria Weighing

AHP has been used in many decision-making projects and suitability analyses. I

adopt it for my green infrastructure suitability analysis, which according to my literature

review is a unique usage.

The main directive guiding the final weighing of the criteria is that technical

requirements are to weigh more than non-technical requirements. The AHP used this

guideline to conduct the pairwise comparison process. Table 4-2 shows the pairwise

comparison matrix, and Table 4-3 presents the normalized pairwise comparison matrix.

These two table contribute to the result of Table 4-4 which indicates criterion weights of

this research. The results of its computation result are represented by the following

formula:

GI suitability = (0.320 * soil) + (0.310 * slope) + (0.177 * imperviousness) + (0.107 * land

ownership) + (0.086 * land use). (4-1)

Table 4-5 proves the consistency of the research. The CI was calculated to be

0.035681, which is less than 0.1, and λ to be 5.142726 which is more than the number

of criteria. That means that the weights for the criteria are consistent and that the result

can, theoretically, be used.

40

Weighted Sum

Following the AHP weighting process, a suitability map was generated using the

weight sum tool in ArcGIS by giving each criterion the value generated for it by AHP.

41

Table 4-1. Suitability Criteria and Ranks

Rank

High Suitability Medium Suitability Low suitability None

Criteria

Soil type Sandy soil Loamy soil Clayey soil /

Percent slope 1-6% 6-12% 12-18% 18% above

Imperviousness 0-100%

impervious

developed

land

20- 49% impervious

developed land

Natural Land

Cover& 0-

20%

impervious

developed

land

Water

Land ownership Public / Private /

Land use type Commercial

, institutional

and open

space

Other Residential

and

conservation

/

42

Table 4-2. Pairwise Comparison Matrix

Soil Slope Imperviousness Land ownership Land Use

Soil 1 1 2 3 4

Slope 1 1 3 2 3

Imperviousness 1/2 1/3 1 2 3

Land ownership 1/3 1/2 1/2 1 1

Land use 1/4 1/3 1/3 1 1

43

Table 4-3. Normalized pairwise comparison matrix

Soil Slope Imperviousness Land

ownership

Land Use

Soil 0.3243243 0.315789 0.292682927 0.333333333 0.333333

Slope 0.3243243 0.315789 0.43902439 0.222222222 0.25

Imperviousness 0.1621622 0.105263 0.146341463 0.222222222 0.25

Land ownership 0.1081081 0.157895 0.073170732 0.111111111 0.083333

Land use 0.0810811 0.105263 0.048780488 0.111111111 0.083333

44

Table 4-4. Criterion Weights

Criterion weight

Soil 0.320

Slope 0.310

Imperviousness 0.177

Land ownership 0.107

Land use 0.086

Table 4-5. Consistency Test Result

Lambda CI

5.142726 0.035681

45

CHAPTER 5 RESULTS

The first five maps show the reclassified criteria and the suitability for each. The

suitability is represented in each cell by the color from the legend on each map.

According to the maps showed below, soil condition in Gainesville is generally

good which can be seen in Figure 5-1. There is only a small percentage of lands with a

low suitability for green infrastructure. Also, the majority of the slope condition of green

infrastructure suitability is fine. The areas with low suitability are mainly water areas

which are excluded from the research. Figure 5-2 displays the percent slope suitability

information. Suitability of land ownership map (Figure 5-3) shows that Gainesville

Renewable Energy Center, University of Florida, Gainesville Regional Airport, Gum

Root Park, Split Rock Conservation area, Bivens Arm Nature Park and other areas with

a color of purple are more suitable for green infrastructures. Figure 5-4 is the land use

suitability map. It indicates that Gainesville Gainesville Renewable Energy Center,

University of Florida, Gainesville Regional Airport, Gum Root Park, Oaks Mall, Split

Rock Conservation area, Bivens Arm Nature Park, car agencies beside NE 39th street

and Main Street, schools in Gainesville and other land with a color of purple are with

high suitability for green infrastructures. Imperviousness map is showed in Figure 5-5. It

delivers the information that Butler plaza, parts of University of Florida, areas around the

intersection of SW Depot Avenue and S Main Street, University of Florida Eastside

Campus, and area around Alley Gatorz Bowling Center beside NE state Rd 24, car

agencies beside NE 39th Street and Main Street, area around the intersection of NW

13th Street and NW 23rd Avenue, area around the intersection of NW 13th Street and

NW 53rd Avenue, part of the Gainesville Regional Airport, part of the Gainesville

http://gainesvillebiomass.com/




46

Renewable Energy Center, Oaks Mall, and other areas representing an orange color

are high suitability areas.

The final map, Figure 5-6, which integrates all five criteria, shows overall green

infrastructure suitability in Gainesville. The suitability is represented by the colors given

in the legend, ranging from blue (low) through yellow (medium) to red (high). The map

can be used as a first step reference or resource for decision makers selecting

implementation sites for green infrastructure. The result shows several potential areas

that are most suitable for green infrastructure implementation. These areas are

Gainesville Renewable Energy Center, Alachua County Corrections Library, Alachua

County Jail, Grace Market Place, car agencies beside NE 39th street and Main Street,

area around Tacachale Center, part of Depot park, 913 SE 5th Street, southern part of

University of Florida, US Post Office on 34th street, Myra Terwilliger Elementary School,

C. W. Norton Elementary School and other areas showing with a red color.



47

Figure 5-1. Reclassified Soil Map

48

Figure 5-2. Reclassified Slope Map

49

Figure 5-3. Reclassified Land Ownership Map

50

Figure 5-4. Reclassified Land Use Map

51

Figure 5-5. Reclassified Imperviousness Map

52

Figure 5-6. Green Infrastructure Suitability Map in Gainesville

53

CHAPTER 6 DISCUSSION

The results of this analysis can be regarded only as a first-step indicator for

decision makers and planners. The areas identified as high suitability are mostly parking

lots. Parking lots are among the most important land cover types for stormwater

management, which shows the model’s feasibility, but this is not the whole story.

A second important point is site specialty. This derives mainly from the fact that

the University of Florida is located within the study area. When a large university is

located in a city, demographic information is different from what it would be in another

city of the same scale. This is one reason I did not consider the minority-and-poverty

factor that was adopted in Walworth Run and Dallas. Poverty data, in which is derived

from demographic data, cannot be used because the large number of students who

would be counted as “low-income” would disturb the results. Another reason for the

importance of site specialty is that most of the soil and slope conditions here are

acceptable for green infrastructure implementation. This fact created ambiguity

throughout the research. However, the two criteria work better at other sites whose soil

and slope condition vary.

A third point is the accuracy of the slope data. These data, as I said, were

generated using the slope tool with the Florida DEM data. When the model was

generated, water areas were all reclassified with a score of –1. The slope generation

process was affected by this reclassification. Normally, water area edges should not be

flat; there should be a slope from the edge of the water area down to the bottom of the

water. Although this did affect the accuracy of the slope data, water areas were

excluded from this research because of the limitations of green infrastructure, as

54

defined here, and the impossibility of implementation. In other words, although the data

are inaccurate in places, this does not hurt the accuracy of the final result.

The fourth matter is criteria selection. A suitability analysis should begin with a

definite goal. In this research, I started by looking at the benefits of green infrastructure

for stormwater management. However, the green infrastructure has considerable

benefits beyond this. The suitability analysis performed in Dallas considered both

stormwater management and the urban heat island effect. In future research, criteria

relating to other goals can be added to further guarantee suitability to green

infrastructures. Criteria selection can also be made on the basis of economic

considerations, such as cost–benefit analyses of specific green infrastructures.

The fifth one is the criteria ranking method. This study uses 1, 2 and 3 to indicate

low, medium and high suitability respectively. It is mainly referred to the study in Dallas

and Town of Berlin. The score of 1, 2 and 3 may affect the accuracy of the result as

some of the attributes in one criteria are given a same score which may be slight

different in suitability. For example, in land use type commercial land and institutional

land are reclassified as high suitability area. However, commercial land may present

higher suitability than institutional land, or green infrastructure functions better in

institutional land than in commercial land. Once there are evidences showing these

differences in suitability for green infrastructure, it can be more accurate to use a larger

scale to distinguish suitability. Further study can be done to detail the differences of

these attribute in each criteria and to differentiate them by using larger scaled numbers.

The sixth one is about the replicability of this study. The method the study is only

referable but not replicable in other place. First, the characteristic of criteria selected

55

must not be the same as what it is in Gainesville. Second, other criteria can be

considered to ameliorate a study based on the goal of a study or the particularity of a

study area.

The final element that should be taken into consideration is the weight given to

each criterion. The core characteristic of a suitability analysis is that it is in fact a

qualitative study represented in a quantitative format. My literature review indicates that

previous studies of green infrastructure suitability, especially in their weighting

processes, did not have statistical support and were conducted empirically. I adopted

AHP in this research, which can alleviate the shortage of weight support for each

criterion. However, the pairwise comparison process in AHP still contains a

transformation step from qualitative to quantitative that lacks solid support. The

comparison result ranging from 1 to 9 can only reduce the error generated by the

transformation but not eliminate it entirely.

56

CHAPTER 7 CONCLUSION

Stormwater management is currently a challenging problem. The implementation

of green infrastructure is one of the most effective ways to conquer this problem. A

green infrastructure suitability analysis can aid in the implementation process. By

considering not only past green infrastructure suitability studies but also research on

land use suitability analyses, and by integrating the data with ArcGIS technology and

the analytical hierarchy process method to provide a solid evidence for criterion weights

and by taking five elements (soil, slope, imperviousness, land ownership, and land use)

into account as suitability determinants, this research has successfully developed a

methodology and mapped out the green infrastructure suitability of the Gainesville area.

57

APPENDIX A DATA PROCESS

Define Study area

1. Add “par_citylm_2011” to the map 2. Select Gainesville city by using select by attribute 3. Export the selected data

The exported Gainesville city will be the study area for this research. All the

analyses are within this area, and the map is named as “Gainesville city limit”

Generate slope percent map

1. Add “FLORIDA DIGITAL ELEVATION MODEL (DEM) MOSAIC - 5-METER CELL SIZE - ELEVATION UNITS METERS” data to the map

2. Use “extract” tool under “surface” in “spatial analyst tools” toolbox to generate Gainesville DEM by setting the mask as “Gainesville city limit”

3. Use “slope” tool under to generate Gainesville slope percent map

Generate soil map

1. Add “SOIL SURVEY GEOGRAPHIC (SSURGO) DATABASE FOR FLORIDA - JUNE 2012” data to the map

2. Convert the vector data to raster by using “feature to raster” under “conversion tools” toolbox and keep “MUNAME” field during the process for future reclassification

3. Use extract” tool under “surface” in “spatial analyst tools” toolbox to generate soil raster map in Gainesville

Generate land ownership map

1. Add “FLORIDA PARCEL DATA - 2012” data to the map

2. Convert the vector data to raster by using “feature to raster” under “conversion tools” toolbox and keep “PUBLICLND” field during the process for future reclassification

3. Use extract” tool under “surface” in “spatial analyst tools” toolbox to generate parcel data in Gainesville

Generate imperviousness map

58

1. Add “NLCD 2011 Land Cover (2011 Edition, amended 2014) - National Geospatial Data Asset (NGDA) Land Use Land Cover” to the map

2. Use “extract” tool under “surface” in “spatial analyst tools” toolbox to generate Gainesville imperviousness map by setting the mask as “Gainesville city limit”

Generate land use map

1. Add “FDOT DISTRICT 2 - GENERALIZED FUTURE LAND USE” data to the map

2. Convert the vector data to raster by using “feature to raster” under “conversion tools” toolbox keep “Descript” field during the process for future analysis

3. Use extract” tool under “surface” in “spatial analyst tools” toolbox to generate land use raster data in Gainesville

59

APPENDIX B RECLASSIFICATION TABLES

Table B-1. Soil Type Reclassification

VALUE COUNT MUNAME New Value

1 96488 MYAKKA SAND 3

2 79875 POMONA SAND, DEPRESSIONAL 3

3 27248 SPARR FINE SAND 3

4 891117 POMONA SAND 3

5 25912 PLUMMER FINE SAND 3

6 63714 POMPANO SAND 3

7 636418 WAUCHULA SAND 3

8 281184 MONTEOCHA LOAMY SAND 2

9 33867 CHIPLEY SAND 3

10 163528 PELHAM SAND 3

11 643017 MILLHOPPER SAND, 0 TO 5 PERCENT SLOPES 3

12 142855 TAVARES SAND, 0 TO 5 PERCENT SLOPES 3

13 151276 SURRENCY SAND 3

14 1672 MASCOTTE, WESCONNETT, AND SURRENCY

SOILS, FLOODED

2

15 92142 NEWNAN SAND 3

16 5280 POTTSBURG SAND 3

17 53793 MULAT SAND 3

18 14278 LOCHLOOSA FINE SAND, 0 TO 2 PERCENT

SLOPES

3

60

Table B-1. Continued


19 7904 PLACID SAND, DEPRESSIONAL 3

20 532204 WAUCHULA-URBAN LAND COMPLEX 2

21 1974 KENDRICK SAND, 5 TO 8 PERCENT SLOPES 3

22 2377 SHENKS MUCK 1

23 80990 SAMSULA MUCK 1

24 132727 KANAPAHA SAND, 0 TO 5 PERCENT SLOPES 3

25 43785 FLORIDANA SAND, DEPRESSIONAL 3

26 93791 WATER 0

27 42925 BLICHTON SAND, 5 TO 8 PERCENT SLOPES 3

28 20107 RIVIERA SAND 3

29 148685 ARREDONDO FINE SAND, 0 TO 5 PERCENT

SLOPES

3

30 21989 PITS AND DUMPS 0

31 13076 APOPKA SAND, 0 TO 5 PERCENT SLOPES 3


33 646323 MILLHOPPER-URBAN LAND COMPLEX, 0 TO 5

PERCENT SLOPES

2

34 37299 MILLHOPPER SAND, 5 TO 8 PERCENT SLOPES 3

35 317397 ARREDONDO-URBAN LAND COMPLEX, 0 TO 5

PERCENT SLOPES

2

61



36 51825 PELHAM, PLUMMER, AND MASCOTTE SOILS,

OCCASIONALLY FLOODED

2

37 14319 ARENTS, 0 TO 5 PERCENT SLOPES 2

38 17547 CANDLER FINE SAND, 0 TO 5 PERCENT

SLOPES

3

39 5499 ZOLFO SAND 3

40 163272 URBAN LAND 2


SLOPES

3

42 105465 BLICHTON-URBAN LAND COMPLEX, 0 TO 5

PERCENT SLOPES

2

43 141033 URBAN LAND-MILLHOPPER COMPLEX 2


SLOPES

2

45 46773 BIVANS SAND, 2 TO 5 PERCENT SLOPES 2

46 5223 FORT MEADE FINE SAND, 0 TO 5 PERCENT

SLOPES

3

47 33458 BONNEAU FINE SAND, 2 TO 5 PERCENT

SLOPES

3

48 11122 BIVANS SAND, 5 TO 8 PERCENT SLOPES 3

49 30037 KENDRICK SAND, 2 TO 5 PERCENT SLOPES 3

62



50 15834 GAINESVILLE SAND, 0 TO 5 PERCENT SLOPES 3

51 2895 PICKNEY SAND, FREQUENTLY FLOODED 3

52 2089 OCILLA, ALAPAHA, AND MANDARIN SOILS,

OCCASIONALLY FLOODED

2

53 3900 STARKE SAND, FREQUENTLY FLOODED 3

54 557 PEDRO FINE SAND, 0 TO 5 PERCENT SLOPES 3

55 8123 NORFOLK LOAMY FINE SAND, 2 TO 5

PERCENT SLOPES

2

56 4755 ARREDONDO FINE SAND, 5 TO 8 PERCENT

SLOPES

2

57 28881 LAKE SAND, 0 TO 5 PERCENT SLOPES 3

58 970 MICANOPY LOAMY FINE SAND, 2 TO 5

PERCENT SLOPES

2

59 942 NORFOLK LOAMY FINE SAND, 5 TO 8

PERCENT SLOPES

2

60 3366 JONESVILLE-CADILLAC-BONNEAU COMPLEX,

0 TO 5 PERCENT SLOPES

2


63

Table B-2. Slope Reclassification

Old Value New Value

0-6 3

6-12 2

12-18 1

18-617.358459 0

64

Table B-3. Imperviousness Reclassification

Value Definition New Value

11 Open Water - All areas of open water, generally with less

than 25% cover or vegetation or soil

0

21 Developed, Open Space - Includes areas with a mixture of

some constructed materials, but mostly vegetation in the

form of lawn grasses. Impervious surfaces account for less

than 20 percent of total cover. These areas most commonly

include large-lot single-family housing units, parks, golf

courses, and vegetation planted in developed settings for

recreation, erosion control, or aesthetic purposes.

1

22 Developed, Low Intensity -Includes areas with a mixture of

constructed materials and vegetation. Impervious surfaces

account for 20-49 percent of total cover. These areas most

commonly include single-family housing units.

2

24 Developed, High Intensity - Includes highly developed areas

where people reside or work in high numbers. Examples

include apartment complexes, row houses and

commercial/industrial. Impervious surfaces account for 80 to

100 percent of the total cover.

3

65



31 Barren Land (Rock/Sand/Clay) - Barren areas of bedrock,

desert pavement, scarps, talus, slides, volcanic material,

glacial debris, sand dunes, strip mines, gravel pits and other

accumulations of earthen material. Generally, vegetation

accounts for less than 15% of total cover.

1

41 Deciduous Forest - Areas dominated by trees generally

greater than 5 meters tall, and greater than 20% of total

vegetation cover. More than 75 percent of the tree species

shed foliage simultaneously in response to seasonal change.

1

42 Evergreen Forest - Areas dominated by trees generally

greater than 5 meters tall, and greater than 20% of total

vegetation cover. More than 75 percent of the tree species

maintain their leaves all year. Canopy is never without green

foliage.

1

43 Mixed Forest - Areas dominated by trees generally greater

than 5 meters tall, and greater than 20% of total vegetation

cover. Neither deciduous nor evergreen species are greater

than 75 percent of total tree cover.

1

66



52 Shrub/Scrub - Areas dominated by shrubs; less than 5

meters tall with shrub canopy typically greater than 20% of

total vegetation. This class includes true shrubs, young trees

in an early successional stage or trees stunted from

environmental conditions.

1

71 Grassland/Herbaceous - Areas dominated by grammanoid

or herbaceous vegetation, generally greater than 80% of

total vegetation. These areas are not subject to intensive

management such as tilling, but can be utilized for grazing.

1

81 Pasture/Hay - Areas of grasses, legumes, or grass-legume

mixtures planted for livestock grazing or the production of

seed or hay crops, typically on a perennial cycle.

Pasture/hay vegetation accounts for greater than 20 percent

of total vegetation.

1

82 Cultivated Crops - Areas used for the production of annual

crops, such as corn, soybeans, vegetables, tobacco, and

cotton, and also perennial woody crops such as orchards

and vineyards. Crop vegetation accounts for greater than 20

percent of total vegetation. This class also includes all land

being actively tilled.

1

67



90 Woody Wetlands - Areas where forest or shrub land

vegetation accounts for greater than 20 percent of vegetative

cover and the soil or substrate is periodically saturated with

or covered with water.

1

95 Emergent Herbaceous Wetlands - Areas where perennial

herbaceous vegetation accounts for greater than 80 percent

of vegetative cover and the soil or substrate is periodically

saturated with or covered with water.

1

68

Table B-4. Land Ownership Reclassification

VALUE COUNT PUBLICLND New Value

1 4261084 1

2 846783 M 3

3 223480 D 3

4 172508 C 3

5 304638 S 3

6 26342 F 3

69

Table B-5. Land Use Type Reclassification

VALUE COUNT DESCRIPT New Value

1 1107338 AGRICULTURAL 2

2 1046153 INSTITUTIONAL 3

3 468155 INDUSTRIAL 2

4 1754622 MEDIUM DENSITY RESIDENTIAL 1

5 204253 COMMERCIAL 3

6 477154 CONSERVATION 1

7 1574 UNKNOWN 2

8 50878 MIXED USE 2

9 576148 HIGH DENSITY RESIDENTIAL 1

10 446 LOW DENSITY RESIDENTIAL 1

11 86822 RECREATION / OPEN SPACE 3

70

LIST OF REFERENCES

Akıncı, H., Özalp, A. Y., & Turgut, B. (2013). Agricultural land use suitability analysis using GIS and AHP technique. Computers and Electronics in Agriculture, 97, 71-82. doi: http://dx.doi.org/10.1016/j.compag.2013.07.006

Alachua County Growth Management Department. (2011). Alachua County

comprehensive plan 2011-2030. Retrieved from https://growth-management.alachuacounty.us/comprehensive_planning/documents/2011_2030_Comprehensive_Plan.pdf

Al-Shalabi, M. A., Mansor, S. B., Ahmed, N. B., & Shiriff, R. (2006). GIS based

multicriteria approaches to housing site suitability assessment. Paper presented at the XXIII FIG Congress, Shaping the Change, Munich, Germany, October.

Banai-Kashani, R. (1989). A new method for site suitability analysis: the analytic

hierarchy process. Environmental management, 13(6), 685-693. Benedict, M., & McMahon, E. T. (2002). Green Infrastructure: Smart Conservation for the

21st Century, 2002a. 1st quote on, 6. Benedict, M. A., & McMahon, E. T. (2012). Green infrastructure: linking landscapes and

communities: Island Press. Berger, D. (2013). A GIS Suitability Analysis of The Potential for Rooftop Agriculture in

New York City. Columbia University. Buchholz, N. (2013). Low-Impact Development and Green Infrastructure

Implementation: Creating a Replicable GIS Suitability Model for Stormwater Management and the Urban Heat Island Effect in Dallas, Texas.

Carver, S. J. (1991). Integrating multi-criteria evaluation with geographical information

systems. International Journal of Geographical Information System, 5(3), 321-339.

Central New York Regional Planning & Development Board. (2012). Green

Infrastructure Planning for Improved Stormwater Management in Central New York. Central New York Regional Planning & Development Board.

City of New Haven. (2012). Green Infrastructure Feasibility Scan for Bridgeport and

New Haven, CT: Evaluation of Green Technologies to Manage Wet Weather Flows.

Collins, M. G., Steiner, F. R., & Rushman, M. J. (2001). Land-use suitability analysis in

the United States: historical development and promising technological achievements. Environmental management, 28(5), 611-621.

http://dx.doi.org/10.1016/j.compag.2013.07.006

https://growth-management.alachuacounty.us/comprehensive_planning/documents/2011_2030_Comprehensive_Plan.pdf





71

Detwiler, Stacey. (2012). Growing Green: How Green Infrastructure Can Improve

Community Livability and Public Health? Retrieved from https://www.americanrivers.org/assets/pdfs/green-infrastructure-docs/growing-green-how-green-infrastructure-can-improve-community-livability.pdf

Duc, T. T. (2006). Using GIS and AHP technique for land-use suitability analysis. Paper

presented at the International symposium on geoinformatics for spatial infrastructure development in earth and allied sciences.

EPA. (2014). Addressing Green Infrastructure Design Challenges in the Pittsburgh

Region. Retrieved from http://water.epa.gov/infrastructure/greeninfrastructure/upload/Pittsburgh-United-Steep-Slopes-508.pdf

EPA. (2011). Evaluation of Urban Soils: Suitability for Green Infrastructure or Urban

Agriculture. Retrieved from http://water.epa.gov/infrastructure/greeninfrastructure/upload/Evaluation-of-Urban-Soils.pdf

EPA. (2010). Green Infrastructure Case Studies: Municipal Policies for Managing

Stormwater with Green Infrastructure. Retrieved from http://www.epa.gov/owow/NPS/lid/gi_case_studies_2010.pdf

EPA. (2014, June 13). What is Green Infrastructure? Retrieved from

http://water.epa.gov/infrastructure/greeninfrastructure/gi_what.cfm EPA. (2007). Reducing Stormwater Costs through Low Impact Development (LID)

Strategies and Practices. Retrieved from http://water.epa.gov/polwaste/green/upload/2008_01_02_NPS_lid_costs07uments_reducingstormwatercosts-2.pdf

Foster, J., Lowe, A., & Winkelman, S. (2011). The value of green infrastructure for urban

climate adaptation. Center for Clean Air Policy, February. Kloss, C. (2008). Green infrastructure for urban stormwater management. Low Impact

Development for Urban Ecosystem and Habitat Protection. Kumar, M., & Shaikh, V. R. (2013). Site suitability analysis for Urban development using

GIS based multicriteria evaluation technique. Journal of the Indian Society of Remote Sensing, 41(2), 417-424.

Malczewski, J. (1999). GIS and multicriteria decision analysis: John Wiley & Sons. Malczewski, J. (2004). GIS-based land-use suitability analysis: a critical overview.

https://www.americanrivers.org/assets/pdfs/green-infrastructure-docs/growing-green-how-green-infrastructure-can-improve-community-livability.pdf

https://www.americanrivers.org/assets/pdfs/green-infrastructure-docs/growing-green-how-green-infrastructure-can-improve-community-livability.pdf

http://water.epa.gov/infrastructure/greeninfrastructure/upload/Pittsburgh-United-Steep-Slopes-508.pdf

http://water.epa.gov/infrastructure/greeninfrastructure/upload/Pittsburgh-United-Steep-Slopes-508.pdf

http://water.epa.gov/infrastructure/greeninfrastructure/upload/Evaluation-of-Urban-Soils.pdf

http://water.epa.gov/infrastructure/greeninfrastructure/upload/Evaluation-of-Urban-Soils.pdf

http://www.epa.gov/owow/NPS/lid/gi_case_studies_2010.pdf

http://water.epa.gov/infrastructure/greeninfrastructure/gi_what.cfm

http://water.epa.gov/polwaste/green/upload/2008_01_02_NPS_lid_costs07uments_reducingstormwatercosts-2.pdf

http://water.epa.gov/polwaste/green/upload/2008_01_02_NPS_lid_costs07uments_reducingstormwatercosts-2.pdf

72

Progress in Planning, 62(1), 3-65. doi: http://dx.doi.org/10.1016/j.progress.2003.09.002

Marney, R. A. (2012). Creation of a GIS Based Model for Determining the Suitability of

Implementing Green Infrastructure: In The Town Of Berlin Maryland. McMahon, E. T., & Clearinghouse, S. W. (2001). Green Infrastructure: Smart

Conservation for the 21st Century: Sprawl Watch Clearinghouse. Mell, I. C. (2010). Green infrastructure: concepts, perceptions and its use in spatial

planning. National Geographic Society. (n.d.). GIS (geographic information system). Retrieved

from http://education.nationalgeographic.com/encyclopedia/geographic-information system-gis/

National Research Council. (2009). Urban Stormwater Management in the United States:

National Academies Press. Northeast Ohio Regional Sewer District. (2011). Walworth Run Green Infrastructure

Feasibility Study. Northeast Ohio Regional Sewer District. Randolph, J. (2004). Environmental land use planning and management: Island Press. Rector, R. P., Duckworth, T., & Obropta C. C. (2012). Rain Gardens and Mosquitoes.

Retrieved from http://www.wrwac.org/fs1175.pdf Rouse, D. C., & Bunster-Ossa, I. F. (2013). Green infrastructure: a landscape approach:

American Planning Association. Saaty, T. L. (2000). Fundamentals of decision making and priority theory with the

analytic hierarchy process (Vol. 6): Rws Publications. USGS. (2008). The National Water-Quality Assessment Program—Informing water-

resource management and protection decisions. Retrieved from http://water.usgs.gov/nawqa/docs/xrel/external.relevance.pdf

Vitousek, P. M., Mooney, H. A., Lubchenco, J., & Melillo, J. M. (1997). Human

domination of Earth's ecosystems. Science, 277(5325), 494-499. WeatherDB. (n.d.). Gainesville, Florida Average Rainfall. Retrieved from


http://dx.doi.org/10.1016/j.progress.2003.09.002

http://education.nationalgeographic.com/encyclopedia/geographic-information%20system-gis/

http://education.nationalgeographic.com/encyclopedia/geographic-information%20system-gis/

http://www.wrwac.org/fs1175.pdf

http://water.usgs.gov/nawqa/docs/xrel/external.relevance.pdf


73

BIOGRAPHICAL SKETCH

Yuxiao Li was born in Xiaogan, China in 1990. He graduated from Southwest

Forestry University in July 2012 with a bachelor’s degree in resources environment

urban rural planning and management. During the 4 years undergraduate study, he had

strong interests in not only planning but also studying abroad. Pursuing the graduate

study in University of Florida, he participated in the International Urbanism Workshop

“Open space System: A Methodology for sustainable Cities”. Also, he worked as an

intern in St. Johns River Water Management District (SJRWMD) to georeference the old

maps.

A GIS-BASED GREEN INFRASTRUCTURE SUITABILTY ANALYSIS …

Documents