Geospatially enabled information systems supporting forest decisions at the Millennium: a U.S. perspective by Jeremy S. Fried USDA Forest Service, Pacific.

Geospatially enabled information Geospatially enabled information systems supporting forest decisions at systems supporting forest decisions at

the Millennium: a U.S. perspectivethe Millennium: a U.S. perspective

by Jeremy S. Fried

USDA Forest Service, Pacific Northwest Research Station

Forest Inventory and Analysis Program

Given 19 May 2000 at Hyytiälä Field Station, Finland

IntroductionIntroduction

My TaskMy Task Current

– No refereed literature; limited gray literature

Status– Relative to whose expectations?

Forest– Include natural resource, chaparral, non-commercial forest?

Information Systems– Only storage/retrieval/display or also models and analysis?– Spatial, quasi-spatial, aspatial?

Expert Opinion SurveyExpert Opinion Survey

20 key informant interviews

Experienced FIS practitioners who

– specify, design, implement and use spatially referenced

information systems intended to support decisions about the

management of forests in the U.S

Snowball sampling

QuestionsQuestions

What does the term “practical forestry information system” mean to you?

Please describe an example from within your organization. – What objectives does that system have? – How does it account for location? – What analysis functions are included (if any) and how are they

implemented? – What inputs are required? – Who are the intended users?

More QuestionsMore Questions

– What has been the biggest barrier to its adoption? To its

effectiveness?

– What do you see as the greatest accomplishment of the system?

How has it changed your organization?

– Did you consider another system before adopting this one?

How did you choose?

– Are you marketing your system to others? If so, which systems

do you see as competitors to yours?

CaveatsCaveats

Inherently biased towards reports of success

Constraints of corporate confidentiality

– Scarce documentation of corporate systems

– Even scarcer reporting in peer-reviewed outlets

Synthesis reflects my interpretations & not necessarily

the USDA Forest Service

Organization of FindingsOrganization of Findings Status of public sector FIS Status of private sector FIS Issues transcending organizational type Examples of interesting FIS cases Barriers to success Accomplishments of FIS Traps to avoid Nuggets of wisdom Discernable trends

U.S. ForestsU.S. Forests

Bureau of Land Management (BLM)Bureau of Land Management (BLM)

Responsibility on 25% of the U.S. land area

Substantial forest lands in the Pacific Northwest

Spatial data involved in 75% of information flow

Government Accounting Office (GAO) ReportsGovernment Accounting Office (GAO) Reports

ALMRSALMRS

Automated Land and Mineral Records System

– Abandoned after 15 years of development and

411 million USD expended

– Lessons learned: lack of management controls and systems

architecture were fatal

Locally in PNW, independently developed FIS are being

used for mapping and to support decisions

ALMRS debacle slowed FIS at NPS and BIA

#

U.S. National Forests

NNational Forests

U.S.

Portland

Legacy FISLegacy FIS

300 forest level FIS constructed over 25 years

– Diverse Purposes

Thematic content

Coding and classification systems

Adequacy of documentation and accuracy standards

– No embedded business rules

– Analysis rarely emphasized

– National system stymied by information infrastructure

Project 615Project 615

Client/server office automation for the USFS– Hardware/software acquisition

– Implementation

– Training

– GIS but one component

Seven year process– Massive, complex, lawsuits by losing bidders

Equipment began arriving in 1995– Vision for agency-wide FIS architecture followed

DatabasesDatabases

Existing vegetation Soil/ecology/geology/geomorphology/climate Aquatic features Fauna Human dimension Constructed/developed features Location information regarding ownership

Current StatusCurrent Status

Local/regional: Zone area province approach– Some success in use of FIS to support

Operational management Planning Outreach

– Continued emphasis on automated mapping and producing maps and data for the public

Ongoing communication among national and regional experts to shape NRIS– Took 3 years to agree on 15 core GIS layers

State FIS ProgressState FIS Progress

50 states with varied missions

– Regulation

– Technical advice and landowner assistance

– Active management of public forests

Thus, different objectives and needs

Washington farther along than most states

Forest IndustryForest Industry

Representing vegetationRepresenting vegetation

Homogeneous stands over time unrealistic

Stands as management and sampling unit

Homogeneity varies by region

Habitat type “stands” supplemented by successional

trajectories

Stands dynamically generated from plot data

Stands/activities as spatial events (in time/space)

Spatial Event ExampleSpatial Event Example

Stand event

Blow-down 1999

Budworm 1998

Clearcut 2000

Legend

A

B

D

C

Spatial-temporal GISSpatial-temporal GIS

Spatial Event SystemSpatial Event System

Offers flexibility to view land from different perspectives

Use of “region” model reduces redundancy

Append instead of replace means information never lost

Business rules and knowledgebase reduce entry/update

time and errors

Role of modelsRole of models

Models planned at design stage– Ensures inclusion of required data– Provides focus for the FIS effort

Most include or link to FORPLAN or other optimizers and schedulers

Several have extensive habitat analysis components

Lessons from Corporate FISLessons from Corporate FIS

Benefits just now beginning, after 15 yearsHigh success depends on

– Genuine and sustained commitment to Full deployment of FIS Using FIS as a basis for decisions

– Visionary leaders or– Visionary technical analyst/catalyst with access to

leadership

Transcendent IssuesTranscendent Issues

Top down or bottom up?Top down or bottom up?

100% bottom up incompatible systems; can’t aggregate to a corporate view

100% top down guarantees consistent systems that receive no local buy-in or local usefulness, and thus usually poor quality data

Mixed approach ?– Energetic/decisive leadership: potential for timely system that

meets many needs– Consensus/committee leadership: painfully slow progress, and

ultimately ?

Evolutionary FIS?Evolutionary FIS?

Most organizations attracted to FIS by desire to automate

map applications

Widely held hope that if you build the database for

mapping, analysis will follow

– This has rarely occurred

– Those who most successfully integrate FIS with their core

business have not followed this path

1 db, 2 db, 3 db, 4…and more1 db, 2 db, 3 db, 4…and more

Tendency to manage FIS components separately, e.g.,

roads/infrastructure, land records, natural resources

– Departmental desire for autonomy?

– Rarely makes using the FIS easier

Critical need: ensuring common database to support

operations and strategic planning

Data qualityData quality

Lineage of pre-existing or legacy data Accuracy specifications often linked to map purposes,

not analytic purposes Analytic limitations due to inadequate accuracy

sometimes not known until analysis (e.g., overlay) propagates errors

Plenty of re-doing in FIS; contributes to time and cost overruns

Melding FIS to the core businessMelding FIS to the core business

Top Priority: analyzing the business process to identify FIS needs

Top Priority: assembling a high quality databaseConflicting view on ordering these priorities

– Build database first, think about analysis later?– Create schematics for analysis, then build database?

Above all, avoid FIS for FIS sake

SoftwareSoftware

No longer matters ESRI products dominate U.S. FIS Oracle database dominates ArcView widely deployed as “front end” Customized A/I and AV interfaces common Custom-built GIS all but extinct Custom-built FIS rarely resold Lots of “general case” decision support tools but few

with broad distribution

Reasons for CooperationReasons for Cooperation

Mixed ownership– Roads– Fire protection

Landscape level processes – Habitat– Hydrologic impacts– Insects and disease infestations

Reduce typical lag of 5-10 years between concept and implementation through partnering

Forms of CooperationForms of Cooperation

Sharing FIS data and costs Interagency vegetation mapping project

– Support Northwest Forest Plan (owls) Mixed grain FIS

– Fine grain on own lands– Coarse grain (RS generated) on others lands

Integrated taxonomic information system (on web) National Biological Information Infrastructure

– Built on FGDC standards User boards/geospatial advisory committees

Data, data, everywhere…Data, data, everywhere…

Almost all federally collected data is free

– Orthophotos

– DEMs

– Digital line graphs

(roads, contours, hydrology on base map)

– Land use/cover

– TIGER (address coded roads and census)

Helps jump start FIS efforts at many organizations

Example 1: National FISExample 1: National FIS

National FISNational FIS Resource Planning Act (RPA) Assessments

– Multi-resource– Status, change and forecasts– Constructed from a national FIS (FIA) and other data

FIA component has– Over 50,000 spatially referenced inventory plots– A nationally standardized plot design– Numerous, precisely measure tree attributes– Plot confidentiality restrictions which make attribute data

unrestricted and location secret to all outside FIA

RPA outputsRPA outputs

Aggregate reporting of status and change for all resources by county, state, region

Projections over time Current and predicted indicators of ecosystem health

– Endangered plants– Endangered animals– Stream flow– Sediment– Habitat structure

What else is FIA FIS good for?What else is FIA FIS good for? Monitoring coarse grain disturbances Numerous aspatial and spatial relationships

– Harvest, owner type, timber price– Forest health, climate event, insect activity– Development, forest attributes, owner type, site– Habitat abundance and distribution

Criteria and indicators computation Carbon flux Invasive exotics “Wall-to-wall” estimates of forest attributes Detailed species maps

Example 2: Collaborative FISExample 2: Collaborative FIS

ARGIS: a participatory FISARGIS: a participatory FIS

Collaborative approach with

– Networked laptops

– Electronic meeting architecture

– Built-in decision support system

– Structured public feedback

Features of Participatory FIS– Query

– Entry

– Annotation

– Relational links

– Simultaneous or anonymous input

– Geographic exploration

– Prioritization

– Database linkage to spatial changes

– Geographic negotiation

– Automation of summary displays for group review and meeting documentation

Example 3: Integrated Example 3: Integrated FIS/Planning SystemsFIS/Planning Systems

EP(x): Ecosystem Planning ExpressEP(x): Ecosystem Planning Express

Began as UC Berkeley research effort under the North Coast pilot and Sierra pilot projects

Morphed to Terra Vision planning system in collaboration with Louisiana Pacific Co. and VESTRA Resources

Evolved to EP(x) and used for sustained yield plans and habitat conservation plans for Louisiana Pacific Co. and Pacific Lumber

Spatiallized FORPLANSpatiallized FORPLAN

EP(x) is a departure from FORPLAN– bigger variety and number of prescriptions

– outputs include ecosystem state variables

– detailed tracking of spatial distribution of ecosystem attributes

– interpretation of ecosystem from perspective of vertebrate wildlife

Based on wildlife habitat relations (WHR)

BarriersBarriers

Takes one year per application for users to come up to speed on the new way of doing things

Hardware and software remain obstacles, especially in agencies (disk storage, software)

Long time and high costs before results are produced can lead to softening of support

Institutional– firewalls, inflexible contracts, strict rules for national consistency

FIS’s Greatest AccomplishmentsFIS’s Greatest Accomplishments

FIS seen as important part of everyone's job

More communication among staff across disciplines and

with public

More thinking from non-traditional perspectives

Users who edit/update feel pride of ownership

Rapid (6 week) FIS update from inventory makes

information current and highly relevant

TrapsTraps

Buying bargain priced FIS technology without projecting

needs and data budget

Individuals driven by personal agendas hijacking the FIS

Reinventing the wheel by forgoing technology because

“not invented here”

AdviceAdvice

Spend time & money early on business analysis and system planning to ensure that system will – Solve real problems – Contain highly accurate spatial and inventory data

Expect data entry/correction to absorb 75% of time Use small extent pilot with full database & analytic functions Plan to update dynamic themes Don’t be afraid to bring in people with skills you need

TrendsTrends

More progress in last 5 years than previous 10

Analysis time frames lengthening

Grain of analysis becoming more fine

More acceptance of remote sensing

Production of wall-to-wall forest coverages

Use of FIS to track activity in development zones

Fuzzing of traditional administrative boundaries

ConclusionsConclusions

Whither Analysis?Whither Analysis?

FIS is more than a data sponge/filing cabinet, but– Biggest use of FIS is still map production

Duerr’s 1979 vintage concept of FIS includes problem ID and prediction capability– Not everyone “get’s it”

System origins limit analytic destiny FIS with strong inventory foundations best positioned to

support analysis

Complex, information rich problems usually

result in simplistic solutions; the more

knowledge we have, the more difficult (not

easier) decisions become.

The Information Fallacy

The Role of FISThe Role of FIS FIS only useful if you

– treat FIS as strategic business resource and– use the information generated to guide management

FIS has no value if decisions are political Some regard federal agencies as reactive and political, so

if– Pre-conceived notion of best outcome, or– Concern that information will lead to lawsuit,– Then increased information reduces manager’s discretion