SEPA
United States
Environmental Protection
Office of Water
(4503F)
. EPA841-R-97-005
May 1997
Designing an
Information Management System
for Watersheds
THE WATEKSBOa* ACADEBW
Information Transfer Series
-------
-------
EPA841-R-97-005
. May 1997
THE WATERSHED ACADEMY
Information Transfer Series, No. 5
Designing an
Information Management System
for Watersheds
Assessment and Watershed Protection Divsion,
Office of Wetlands, Oceans nd Watersheds (4503F)
and
Permits Division,
Office of Wastewater Management (4203)
Office of Water
U.S. Environmental Protection Agency
401 M Street, SW
Washington, DC 20460
-------
No. 5
Designing an Information Management System for Watersheds
This document was prepared for the U.S. tn^ronme^ ^^^^ ^be^
Greg' Currey^Tthe EPA Office of Water are the project managers.
Notice
This document has been subjected to U.S. Environmental Protection Agency review and
tasbeen^^proved for publication. Publication does not signify that the contents necessariy
mflecTthe viewsand policies of the Environmental Protection Agency or of any other
organization represented in this document. Mention of trade names or commerc.al products
does not constitute endorsement or recommendation for use.
This report should be cited as:
U.S. Environmental Protection Agency. 1997. Designing »n inforrnator' ^emerrt
system for watersheds. EPA841 -R-97-005. Office of Water (4503F), United States Envi
ronmental Protection Agency, Washington, DC.
To obtain a copy free of charge, contact:
National Center for Environmental Publications and Information (NCEPI)
Phone:(513)489-8190
Fax: (513)489-8695
This EPA report may also be available on the Internet for browsing or download at:
http://www.epa.gov/OWOW/watershed/wacademy/its.html
-------
_, . . Information Transfer Series
Watershed Academy
Foreword
The watershed approach has changed the way that the U.S. Environmental Protection
Aaencv (EPA) and other federal, tribal and state agencies formerly managed water re-
sources programs. We now generally recognize that the critical environmental .ssues
facing society are so intertwined that a comprehensive, ecosystem-based and commu-
nity-based approach is needed. We also recognize that solving environmental problems
depends increasingly on local governments and local citizens. Thus, the need to integrate
acmsstrSna. water program areas (e.g., flood control wast^^^^^
source pollution control) and to cooperate across levels.of government (federal^ state,
tribal, local) and across public and private sectors is leading toward a watershed ap-
proach. .
Public and private organizations, academic institutions, and citizens and the.r govern-
ments in thousands of communities across the nation are forming partnerships and learn-
q new ways to manage their watersheds together. These groups seek guidance and
examples of watershed approach success stories after which to model their own act.v,-
ties The EPAOffice of Water established the Watershed Academy to help address these
needs by providing training for watershed managers based on local state, tr.bal and
federal experiences in implementing watershed approaches throughout the past decade.
The Watershed Academy provides technical watershed information and\ ^^
through live training courses, the Internet, and published documents The Academy of-
fe s live training courses on the basics of watershed management and maintains a tra.n-
ina cSSloaue concerning where to obtain more advanced training. An Internet d.stance
earnfng program cale39Academy 2000 is being developed to help serve the trammg
neeSs ofPthose who cannot attend the live courses. ^Watershed Academy also prc-
vides watershed approach reference materials, such as th.s document, through the Wa-
tershed Academy Information Transfer Series.
This document, number 5 in the Series, is an introduction to theh^^;|^-
ment responsibilities and challenges facing any watershed group. The d?M£J
£e fundamentals of identifying information management need s,^^^^
bases, evaluating hardware and software options, and developing implementation plans.
The Information Transfer Series titles include:
no. 1: Watershed protection: a project focus (EPA841-R-95-003)
no 2- Watershed protection: a statewide approach (EPA841-R-95-004)
no 3- Monitoring consortiums: A cost-effective means to enhancing watershed
data collection and analysis (EPA841-R-97-006)
no 4: Land cover digital data directory for the United States (EPA841 -B-97-005)
no 5: Designing an information management system for watersheds
(EPA841-R-97-005)
no. 6: Information management for the watershed approach in the Pacific North-
west (EPA841 -R-97-004)
no. 7: Watershed Academy catalogue of watershed training opportunities
-(EPA841-D-97-001)
-------
-------
.... , . AMflAm.. Information Transfer Series
Watershed A cademy
Abstract
To build a foundation for sound decision-making and reduce electronic data transfer
time it is important to build an information management system concurrent with the
development of a watershed management framework. Based on watershed management
experience in several states, 7 key steps have been identified to design an information
management implementation plan.
1) Establish information management design and implementation team
2) survey watershed planning partners
3) identify and prioritize data needs
4) integrate/relate existing databases and develop new databases
5) evaluate hardware and software configurations
6) evaluate organizational, staffing, and support issues
7) develop short-and long-range implementation plans.
Recommended milestones and guiding policies for each of these steps are detailed in
the report.
Keywords .
data consortium
information management system
CIS packages
information database
-------
-------
Watershed *~*i Information Transfer Series
Contents
Disclaimer
Foreward '
... v
Abstract '"
- ' vii
Contents
vii
List of Figures ..: ""
Introduction
A Seven-Step Process for Designing an Information Management System
Step 1: Establish a design and implementation team 3
Step 2: Survey watershed .planning partners 3
Step 3: Identify and prioritize data needs 4
Step 4: Integrate/relate existing databases and develop new databases 5
Step 5: Evaluate hardware and software configurations 7
Step 6: Evaluation organizational, staffing, and support issues 12
Step 7: Develop a short- and long-term implementation plan ...14
Recommended Milestones and Guiding Policies .15
Appendix A: Definitions for Six Basic Computer System Components -.19
Figures
Exhibit 1. Steps for Designing an Information System 1
Exhibit 2. How Information Management Relates to a Watershed Management
Cycle '-.' 2
Exhibit 3. Comparison of Three User-Friendly GIS Packages 8
Exhibit 4. Comparing ARC/INFO and ArcView2 (1996) --10
Exhibit 5. Option A-PC ArcView2 11
Exhibit 6. Option B - ArcView2 X-Emulation (or X Terminal) 12
Exhibit 7,Option C-ArcView 2 plus ARC/INFO --13
Exhibits. Recommended Milestones and Guiding Policies 15
VII
-------
-------
Watershed Academy
Information Transfer Series
Introduction
Drawing from states' watershed management experiences, this report describes the
design and development of an effective watershed management information system in a
7-step process (Exhibit 1). Although the primary audience is states, the process outlined
can also help local governments, watershed associations, and others interested in design-
ing an information system to support watershed management. Integrating and assessing
information from diverse programs, agencies, and institutions are essential in each stage
of watershed management (see Exhibit 2). This management approach begins with pub lie
outreach, moves through various stages of data gathering and interpretation, and finally
maps management strategies in a watershed action plan.
Time invested building an information management system concurrent with develop-
ing the watershed management approach yields long-term returns. It provides a strong
foundation for management decisions, and reduces time spent collecting data and con-
verting from one computer format or language to another.
One product of this 7-step process for information management system design is an
information management, implementation plan. When planning, it is important to remem-
ber that full implementation is a long-term commitment, and will likely take years. Expe-
rience shows that a successful program requires that staff be practical and flexible in
meeting short-term needs, and at the same time, be dedicated to its long-term vision.
1
1
Step 3. Identify and Prioritize Data Needs
Step 4. Integrate/Relate Existing
Step 5. Evaluate Hardware and
Databases and Develop New Databases
i
Software Configurations
1
Step 6. Evaluate Organizational
, Staffing, and Support Issues
1
1
Exhibit 1. Steps for Designing an Information System
7
-------
Designing an Information Management System for Watersheds
Review/Assess Watershed Information
Overlay unit coverage and potential stressors
Relate stress coverage to stream stretch coverage
. Relate stress coverage to stream classification and use attainment
Compile/convert text files
Evaluate/Present Information
Trend analysis and statistical summaries
Summaries of basin priorities per water quality assessment
reports
L CRP contractor guidance
Prioritize Watershed Issues
. Document/manage feedback from planning stakeholders .
Maintain Matrix of priority issues and watersheds
Formalize Monitoring and Data Collection Partnerships
Establish and document design and QA/QC protocols
Identify Baseline and Strategic Sampling Locations
» * Digitize and create coverages for GIS
V. .^____^__^
Metadata
Develop database of existing databases
Develop bibliography of relevant studies, reports, etc.
Compile Relevant Water Quality Issues
Convert and digitize as needed
QA/QC and store data
Manage Data
Create databases
Import historical data and input new data
QA/QC
'Assess Priority Watershed Issues
Reconfigure databases for model input-sorting, relating,
retrieving
Use data to develop/apply models to further quantify
relationships and test future conditions
Generate screen displays, maps, tables, charts
. Delineate geographic areas for strategy development
/*~"'.
Develop Strategies
* Identify specific goals and objectives
Identify resource management options and means for
implementation .
. Test management options in models and evaluate effectiveness
Develop matrix/rank options per goals, objectives, and cost
effectiveness
Prepare Draft Plan
Assemble recommendations, management strategies, schedules,
maps, and funding mechanisms from stakeholders
. Summarize information from, previous steps
'Draft Plan Review/Hearings
Document/manage written and oral feedback from stakeholders
Revise plan (maps, text files, etc.)
Conduct Public Outreach
and Education
Establish Basin Goals and Priorities and
Develop Monitoring Plan
Implement Strategic Data Collection and\
Monitoring Plans )
Compile and Maintain Detailed
Information and Data
Analyze and Evaluate
Information and Data
Quantify Impacts and Resources J
and Rank Watersheds ]
Develop Management Strategies and
Priority Watersheds
Document Management Strategies J
and Recommendations J
Finalize Watershed Action Plans J
i^8!WWff«"5*"^!?55?^r'!^"^^^^^^
Implement Watershed
Management Strategies
Exhibit 2. How Information Management Relates to a Watershed Management Cycle
-------
Watershed Academy_ ' , Information Transfer Series
A Seven-Step Process for Designing an
Information Management System
Step 1. Establish an Information Management Design
and Implementation Team
A team involved in all phases of design and implementation should conduct the infor-
mation management planning process. Generally, a team should include representatives
from key stakeholder groups and staff knowledgeable about information systems. The
design and implementation team will likely evolve into a data consortium that supports
information needs throughout the initial and subsequent watershed management cycles.
Selection of team members should consider where program and agency coordination
will be crucial to the success of watershed management and also individual knowledge of
database management technology and geographic information system (CIS). The team
should have clearly defined goals and tasks and a mechanism for providing feedback to
the watershed management group. The management group can provide additional guid-
ance for the team's efforts by developing policies (e.g., level of program interconnection
desired, cost constraints, etc.) or general conceptual models of database needs.
Once the team is established, it should conduct a self-assessment to answer ques-
tions such as:
What is each member's skill or expertise?
Are there gaps? If so, do you need an information management consultant?
What else is happening that may affect our design efforts?
After this self assessment, the team is ready to take the next steps in designing an
information system.
Step 2. Survey Watershed Planning Partners
Your water resources program probably uses information from diverse sources in a
myriad of electronic formats. Many state programs have general databases to support
their programs in addition to a GIS. Staff use these general databases to analyze trends,
assess and summarize current conditions, and generate reports-tasks that require such
software functions as storing information, analyzing statistics, sorting and extracting
data, converting data, creating summary charts, and word processing. GIS, on the other
hand, is designed for those applications with geographical references.
-------
Designing an Information Management System for Watersheds
Step 2: Key Survey Questions
What are your group's primary responsi-
bilities related to watershed manage-
ment?
What databases, maps, and geographic
records do you currently use? How do
you use or analyze them? Which are im-
portant for watershed management?
What is their current format?
What map-related problems (e.g., scale
and accuracy) does your group have?
What priority would you give each prob-
lem?
In the watershed management context,
what mapping, data management, and
analysis would you like to accomplish
with a CIS? What functions should the
GIS perform? What non-GIS functions
need to be performed?
What are the obstacles to developing an
information system for the watershed
management approach? How could they
be overcome?
What potential sources of funding and
staff support can your group provide?
What projects might impact implementa-
tion of an information system to support
watershed management?
Designing an information system to support
watershed management usually requires that
managers identify specific requirements through
a user survey. The survey can target a wide
range of divisions, departments, and agencies,
or can target representatives on the information
system design team. In any case, the survey
should include existing and potential watershed
management partners that will use or contrib-
ute program data.
Many state and regional agencies have de-
signed user surveys that can be drawn upon.1
States that conducted user surveys to help de-
velop an implementation plan offer the follow-
ing advice:
Get people thinking geographically in day-
to-day business before designing specific
analytical programs or macros.
Initially focus on overlay/watershed ori-
entation and mapping functions for the
GIS.
Delay development of expensive, custom-
ized GIS/database application and assess-
ment tools.
Carefully set priorities on data essential
for watershed management.
Staff will have a better sense of what they really need when they think geographically,
know tL time involved in using GIS, and understand the constraints of develop.no or
acquiring certain data layers.
» ' , '
Step 3. Identify and Prioritize Data Needs
If you are striving for rapid implementation of watershed management, andI cost effec-
tiveness using survey findings to distinguish essential mapping and assessment data
f om second-priority and low-priority data is important. Although user surveys often
generate lengthy lists of data needs, states have found that some base mapping cover-
ages and analytical data are most frequently needed for watershed management.
1993. GIS Conceptual Design for the North Carolina Water Quality
' Regional Commission. 1990. Implementation Plan for a Regional Geo-
graphic Information System.
-------
Watershed Academy
Information Transfer Series
Geographic management units (e.g., river basin/watershed boundaries, and
subwatershed boundaries), major hydrology, major roads, and political boundaries
(e.g., county and municipal)
Sub-basin units, ambient monitoring sites, National Pollutant Discharge Elimination
System (NPDES) permittee outfalls, US Geological Survey (USGS) flow gauging
sites, stream classifications, stream use information, water supply intake loca-
tions, regulated flow structures, primary shellfish waters, fish nurseries, closed
shellfish waters, superfund sites, landfills, wetlands, land use/land cover, small
area population/household estimates and projections2, slopes, soils, and stormwater
management
Data priorities will vary with the scope of watershed management.
After developing data priorities, develop an
action plan and timeline for acquiring essential
data, starting with the initial target basins. Make
the most of existing data, even though they may
not be perfect or complete, and improve data
quality over the duration of the implementation
plan. Delaying your watershed management pro-
gram for 5 or more years while you develop high-
quality data for each category is both impractical
and unnecessary.
Step 4. Integrate/Relate
Existing Databases
and Develop New
Databases
Formalizing a data consortium is an impor-
tant first step in integrating and relating existing
databases. A consortium maximizes ihteragency
talents, expertise, and resources; reduces data
transfer and conversion headaches; and pro-
duces higher-quality data that can lead to higher-
quality analyses. Documenting roles,
responsibilities, and procedures extends institu-
tional memory, which is especially important
during times when staff changes are common.
The following milestones are suggested for for-
malizing the data consortium:
Step 4: Key Questions Related to
Existing and New Databases
How often will you create additional GIS
coverages upon completion of base wa-
tershed management coverages?
How often will you update watershed
management data files and existing
GIS map coverages?
What map coverages are, currently used
(such as NPDES sites, landfills, and
Superfund sites)? What databases will be
related to map coverages (such as effluent
water quality information related to an
NPDES site file or stream classification re-
lated to stream stretch file)?
What databases will not be linked or re-
lated to GIS?
How will you relate water quality and sup-
ply data to specific streams, stream, seg-
ments, and aquifers?
What is the smallest management unit you
should have, remembering that, nesting up
is easier and more accurate than nesting
down?
How will you relate important off-stream
watershed data (such as location of land-
fills)?
Who will be-the custodian of watershed
management data?
2 Land use, wetlands, and population/household projections for small areas are the most difficult data to acquire at
the high spatial resolution and accuracy needed for small areas.
-------
No 5 Designing an Information Management System for Watersheds
Develop data transfer and quality assurance/quality control (QA/QC) protocols
Evaluate data sources, including software and language compatibility
Choose key relational fields and unique identifiers
Develop a detailed plan for converting, relating, integrating, and updating data-
bases
The fundamental question to answer in designing your database system is, "How
connected and how centralized should it be?" Consider strengths and potential pitfalls of
the following two database management models which are at different extremes in terms
of connectivity and centralization. Model 1 is decentralized, but connected electronically
Model 2 is a centralized, relational database. For each model, we have assumed that the
state has created one or more watershed management coordinator positions and has
developed a CIS program thatcould support multiple agencies.
Model 1: Watershed management coordinators would use the state's CIS program
to compile existing CIS databases and develop new ones; they would use EPA s
Reach File 3 (RF3), and other database and spreadsheet files provided by other
branches, departments, and agencies. These partners would be the responsible
custodians of their datasets, making sure that watershed coordinators receive
updated files. Additionally, different divisions or departments would write chap-
ters or sections of the watershed management plan and provide the narrative and
tables to the basin coordinators in electronic format. Partners exchanging large
files are connected via Internet; other file exchanges are made through e-mail or
tapes and disks.
The strengths of Model 1 include the use of wide resources and staff who have
incentives and insights into data QA/QC. Potential pitfalls include the risk of nu-
merous, incompatible types of databases; lack of standardized file conversion and
file transfer procedures, which creates inefficient and frustrating data exchange;
and questionable data due to lack of overall QA/QC procedures. These pitfalls
could be addressed through standard operating procedures that outline QA/QC,
procedures and compatible database structures.
Model 2- Watershed management coordinators use one relational database, de-
velop and update most watershed databases in-house, and create CIS coverages
and conduct spatial analyses in-house, and write the watershed, management plan
in-house. Partners transferring data are required to use uniform software and data-
base structures; databases or files retrieved from partners are linked to the central-
ized database.
Strengths of Model 2 include more control over QA/QC, data documentation, and
other procedures. Potential pitfalls include large demands on watershed manage-
ment planning coordinators. Custodians of data and information are not as fam.l.ar
with specific databases as program staff. Also/on-site data storage requirements
are great, and the larger and more complex the relational database, the more likely
that it will fail.
-------
, , . . Information Transfer Series
Watershed Academy __; . -
After your watershed management team has reviewed and discussed two or three
conceptual models, it should develop guidance for the information system design team.
Some decision points include:
Develop goals consistent with the long-term vision for the watershed management
information system;
Develop criteria for evaluating model options;
Develop a consensus conceptual model; and
Identify key short-term steps to transfer, relate, and convert data for the initial
watershed plan.
Step 5. Evaluate Hardware and Software Configurations
In evaluating hardware and software configurations, it is important to understand
basic system components and some frequently used terms. These are provided in Appen-
dix A.
Geographic Information Systems: A General Overview
Evaluating existing and future conditions in subwatershed units and targeting high-
priority problem areas, are central to the watershed management approach. These efforts
not only require compiling and storing tremendous volumes of information, but in many
cases, involve referencing the data geographically, overlaying georeferenced data for an
entire river basin, then overlaying sub-basin boundaries for detailed analysis. Analyses are
conducted at the sub-basin level and then aggregated to the basin level to see the overall
picture.
Most states use a CIS to "georeference" data types or layers, overlay layers, conduct
spatial analysis, and generate report maps and charts. Exhibit a compares ArcV,ew2
Maplnfo, and Atlas GIS-three popular, user-friendly G1S packages. Technology changes
quickly; therefore, you will need to refer to trade magazines and vendors for up-to-date
comparisons.
General Comparison of Arc/Info and ArcView2
To a large degree, the decision to use a specific CIS software package should be
based on the volume of information and the complexity of the operations that are ex-
pected, as well as the skills of the users. Will users require a large, regional-scale data-
base that supports heavy use, or is the need restricted to a sing e local-scale watershed?
Can the team afford a full-time CIS specialist, or will a small, m.n.mally funded staff need
to develop basic GIS survival skills?
As GIS software has diversified, it has become possible to trade off complexity and
power for simplicity and a higher degree of user-friendliness. A Qood exampte of the
opportunity for this kind of tradeoff is illustrated by the choice between ARC/INFO, which
-------
Ato. 5
Designing an Information Management System for Watersheds
Exhibit 3. Comparison of Three User-Friendly GIS Packages"
8
-------
Watershed Academy Information Transfer Series
is a powerfufbut very complex UNIX-based GIS, and ArcView, which is a more easily
learned but less powerful, PC-based or UNIX-based package from the same vendor.
Exhibit 4 compares these software packages in terms of the information management
functions each will support. Again, for up-to-date comparisons, it is recommended that
you consult trade articles and vendors.
Evaluating Three General Hardware and Software Configurations
Step 4, Designing a Database Management Model, reviewed conceptual models for
consideration. A similar exercise is helpful in evaluating hardware and software configu-
rations The hardware and software configurations outlined below fall at different points
on the continuum, from centralized to decentralized, and electronically connected to
disconnected. Of course, options and capabilities will change over time.
Option A
In Option A, watershed management coordinators use PC ArcView2 , numerous data-
bases assembled from basin planning partners, and Internet connections for large file
transfers (Exhibit 5). This option is decentralized and adequate if the majority of basin
data and GIS functions will be overlaying, selecting and viewing coverages, and design-
ing maps and charts. Option A also requires that coordinators have a good working
relationship with outside GIS partners to create and update map coverages, generate
color maps, and conduct GIS spatial analysis. It also assumes that there are Internet
connections between major data partners.
The limitations of Option A include a five-step conversion process to export ARC/INFO
files into PC ArcView2. Also, agencies must store both UNIX and DOS copies of each file
(which increases the complexity and cost of file management) and must install additional
software to enable ethernet or network communication between DOS and UNIX. Finally,
Option A is operationally slower than other options.
Option B
In Option B watershed coordinators use ArcView2 run from a UNIX workstation through
PC X-emulation or X-terminals, databases retrieved from basin planning partners, and
Internet connections for large file transfers (Exhibit 6). Option B is suitable when the
majority of basin data and GIS functions will be overlaying, selecting, and viewing cover-
ages, and designing maps and charts. Option B also requires that the coordinators have a
good working relationship with outside GIS partners to create and update map coverages,
generate color maps, and conduct GIS spatial analysis.
The main limitation of Option B occurs if you do not already have a workstation, in
which case this option will require additional hardware purchase and staff training on the
UNIX system.
Option C
In Option C, watershed management coordinators would run both ARC/INFO and
ArcView2 from a UNIX workstation through X-emulation or X-terminals to create, up-
date, and store most basin map coverages and databases in-house. Option C is best if the
-------
Designing an Information Management System for Watersheds
What Both ARC/INFO &
ArcView2 Can Do
Identify, List: Choosing sites, displaying and
listing information about them
Select, Re-select: Selecting sites that meet
certain criteria in the relational database, dis-
playing and listing information about them
Display overlay: Overlaying map features to
show their relationship on screen or print out
Buffer Display: For example, showing the
land area, streams, landfills, etc. that fall
within a prescribed radius or linear distance
Import dBASE, Paradox, and other database
files related to map coverages
Update information/tabular databases related
to map coverages
Map design and generation
Summary charts and tables
Additional ARC/INFO Functions
Overlay existing coverages (through joining,
intersecting, selecting, etc.) to create and dis-
play new coverages
Overlay coverages for calculation, analysis,
and display (for example, calculating the to-
tal number of housing units and average hous-
ing density per Census block group in a given
watershed management unit)
Buffer coverages for calculation, analysis, dis-
play (for example, calculating the number of
stream miles or land area within a certain
linear distance or radius)
Perform special functions such as network-
ing and address matching
Digitize new coverages
Differences Between
ArcView2 & PC ArcView2
ArcView2 and ARC/INFO are run in a UNIX
operating system environment, generally run-
ning off a UNIX workstation [Through your
network and/or special software (X-emula-
tion), you can link a PC or X Terminal to the
UNIX Workstation, essentially running ARC/
INFO or ArcView2 off the workstation while
entering commands and viewing the results
at your PC or X Terminal.]
PC ArcView2 is generally run in a DOS envi-
ronment.
Functionally, there is no difference between
PC ArcView2 and ArcView2 run through X-
emulation, due to the differences between
the UNIX and DOS systems, however, some
key network, file management, and process-
ing differences need to be considered.
For instance, with PC ArcView2 running in a
DOS environment on a 486:
A five-step, export-import conversion pro-
cess is necessary to use map coverages
or databases created in ARC/INFO (in the
UNIX environment) (No translation/con-
version is needed for ArcView2 using X-
emulation on a PC or using an X-terminal).
Must keep two updated copies of every
file (one in DOS, one in UNIX-based
environment), which makes managing/
storing files more complicated and expen-
sive.
- Running large files off a PC is much slower
than off the workstation (directly or
through X-emulation).
It is necessary to have additional ethernet
or network software, such as PCNFS, that
allows communication between the UNIX
and DOS operating systems.
- Fewer digits are allowed in PC ArcView2
names than in ARC/INFO and ArcView2,
which means database structures and no-
menclature would need to be changed in
ARC/INFO before converting a file to PC
ArcView2a process that can be very
time-consuming.
Exhibit 4: Comparing ARC/INFO and ArcView2 (1996)
10
-------
Watershed Academy
Information Transfer Series
Watershed Management Coordinators
PC
- PC ArcView2
- Various databases
Paradox
dBase
- Text files and wordprocessing
File Storage - ready access
Post Script Laser Printer for Report
Maps
Network Fileserver/Intemet Connection
External Partners*
UNIX Workstations
- ARC/INFO
Pen Plotter and Electrostatic Printer for
Color Presentation Maps .
File Storage - ready access & large
archives storage
PC
- Various databases
Paradox
dBase
- Text files
HardcopyMaps '
Network Fileserver/Internet
Connections
Assume GIS partner provide GIS coverages and other data;
other partners provide data and text files and maps ,
Adequate if: Majority of watershed planning data/GIS functions are viewing, overlaying, selecting
coverages and designing/generating maps and charts.
Most GIS map coverages are static and you only need to regularly update associated
databases. ,
Good working relationship with a GIS/Spatial Analysis Center in creating and
updating map coverages and conducting special analysis.
Internet connections between major partners for transferring large files.
Limitations: 5-step conversion process to export ARC/INFO into PC ArcView2
Copy of UNIX and DOS file increases complexity and cost of managing tiles
Need additional software to allow ethernet to communicate between DOS and UNIX
Exhibit 5. Option A - PC ArcView2
watershed management program needs the full range of GIS functions
including creating GIS coverages, and conducting special analyses, or if uniform proce-
dures and database structures are needed (Exhibit 7).
The limitations of Option C include higher investment in hardware and software,
basin planning staff and training, duplication in investment in basin partners ex-
pertise, and GIS technology.
After reviewing and discussing two or three conceptual models, the watershed man-
agement team should develop specific guidance for the information system design team.
Step 5 decision points include:
Develop long-term goals.
Select criteria for evaluating options.
11
-------
No. S
Designing an Information Management System for Watersheds
Watershed Management Coordinators |
UNIX Workstation
- ArcView2
File Storage - ready access and
archives
X Terminal
- X-emulation ArcView2
PC
- X-emulation ArcView2
- Various databases
Paradox
dBase
- Text files and wordprocessing
Postscript Laser Printer for Report
Maps
External Partners*
UNIX Workstation
- ARC/INFO
Pen Plotter and Electrostatic Printer
for Color Presentation Maps
File Storage - ready access & large
archives storage
PC
- Various databases
Paradox
dBase
- Text files
Hardcopy Maps
Network Fileserver/Internet
Connections
Assume GIS partner provide GIS coverages and other
data; other partners provide data and text files and maps
Very Strong if: Majority of basin planning data/GIS functions are viewing, overlaying, selecting
coverages and designing/generating maps and charts.
Most GIS map coverages are static and you only need to regularly update
associated databases. . .
Good working relationship with a GIS/Spatial Analysis Center in creating and
updating map coverages and conducting special analysis.
Internet connections between major partners for transferring large files.
Limitations: Requires additional hardware (workstations) and staff training.
Exhibit 6. Option B - ArcView2 X-Emulation (or X Terminal)
Design a consensus conceptual model.
Identify key short-term steps that support initial basin plans.
Step 6. Evaluate Organizational, Staffing, and Support
Issues
Whether you hire new staff, reassign existing staff, or some combination, below are
some responsibilities that should be addressed to have a smoothly running system.
Watershed Management Coordinator: As watershed plans are developed the Wa-
tershed Management Coordinator arranges meetings with Partners and stakehold-
ers; communicates data and information needs; tracks progress of data acqu^on
and tasks related to plan development and updates; encourages and assists with
the integration of data and results across branches, d.v.s.ons departments and
agencies; communicates and arranges meetings with stakeholders; assembles and
12
-------
Watershed Academy
Information Transfer Series
Watershed Management Coordinators
UNIX Workstation
- ARC/INFO
- ArcView2
File Storage - ready access and
archives .. -
Digitizing Table
X Terminal
- X-emulation ARC/INFO and
ArcView2
PC
- X-emulation ARC/INFO and
ArcView2
- one relational database (Paradox)
- Text files and wordprocessing
External Partners*
PC
- Various databases
Paradox
dBase
- Text files and word processing
Hardcopy maps
Network Fileserver/Internet
Connections
Assume external partners would provide limited
databases and text files.
Very Strong if: Watershed management program needs full ranges of GIS functions in house
including creating coverages and conducting special analysis.
Limitations:
Much higher investment in additional hardware/software.
Much higher investment in additional staff and staff training.
Duplicates watershed management partners' expertise and investment in GIS
technology. \
Exhibit 7. Option C - ArcView2 plus ARC/INFO
assesses comments from stakeholders; and maintains a bibliography of basin-
related documents and a library index of related data frtes. The coordinator a^o
helps summarize and assess data and narrative contributed by watershed manage-
ment partners.
Technical Coordinator: The Technical Coor-
dinator conveys the overall vision of water-
shed management technical functions;
serves as liaison with outside agencies and
partners; administers overall equipment,
staffing, training and purchasing.
Systems Administrator: The Systems Ad-
ministrator configures, optimizes, and main-
tains the GIS network, peripherals, software,
and databases. Administration includes soft-
ware installation, operating system up-
grades, hardware installation, hardware and
software troubleshooting, system back-ups,
and security.
Step 6: Key Questions Related
to Organizational, Staffing, and
Support Issues
Which responsibilities should be
centralized versus decentralized?
Should we hire new staff or allo-
cate additional information man-
agement responsibilities to existing
staff?
Who takes the lead in marshaling
resources and managing informa-
tion?
-------
A/o. 5
Designing an Information Management System for Watersheds
Database Administrator; The Database Administrator manages the creation and
maintenance of CIS and other databases, defines and enforces QA/QC proce-
dures, structures data sources, identifies data sources, coordinates data transla-
tion and exchange activities, and manages access to CIS watershed management
databases.
GIS Technician: The CIS Technician enters data, manages the database, creates
maps, generates reports, performs high level GIS functions; hardware and soft-
ware troubleshooting.
GIS Analyst/Programmer: This GIS specialist performs GIS application analysis,
design, coding, training, and technical support as well as customizes software for
watershed management applications.
Lessons Learned from Other States
Top priority: invest in watershed management coordinator(s) and GIS/database
technician. ,
An agency should not assign watershed management information tasks to existing
staff on top of their existing workload and expect to meet schedules; watershed
management staffing needs should be included explicitly in the organizational bud-
get and workplans.
These days, state and other agencies are usually facing no increase in, and often a
decrease in resources. Therefore, implementing watershed management means
making tough choices about allocating existing staff time to watershed manage-
ment and decreasing staff time in other activities.
Across the country, states have successfully convened watershed management
partners to negotiate a management framework that enables them to 1) get formal
buy-in from partners on when and how they will contribute to watershed manage-
ment; 2) allocate program resources more flexibly and cost effectively; and 3)
leverage outside resources for information management and other activities.
Step 7. Develop A Short- and Long-Range
Implementation Plan with Realistic Funding for
Staff, Hardware, and Software
Step 7 essentially pulls together the work of Steps 1-6 into a multi-phase, 5-year
information management system implementation plan. Key components of the plan in-
clude:
staffing (including training)
hardware
software
14
-------
... J . . .. j.,, Information Transfer Series
Watershed A cademy
Step 1. Set Up an Information Management Design and Implementation Team I
. Establish clear goals, tasks, and schedules as well as a mechanisnffor communicating with
the watershed planning workgroup
. Inventory skills and knowledge of the team and bring in other expertise as needed
Inventory existing information systems to be linked to watershed planning
Identify potential users of the information system
Step 2. Survey Watershed Planning Partners
For each group of users, identify primary responsibilities related to watershed planning/
management
Document relevant existing databases, maps, and geographic records and their formats
Identify data gaps
Distinguish CIS and non-GIS mapping, data management, and analytical functions
Identify current obstacles to developing an information system and potential solutions
- Identify potential sources of funding and staff support
Identify current or planned projects, that could impact implementation of an information
system
Step 3. Prioritize Data Needed
Prioritize data need
Create a schedule for developing or acquiring data
Step 4. Integrate/Relate Existing Data and Develop New Data
Develop and formalize data transfer standards and QA/QC protocols
Develop and formalize a plan for transferring, relating, integrating, and updating data
Evaluate data sources (including quality and compatibility)
Choose key database relational fields for geographic analysis
Determine how database relational fields will be linked to CIS hydrology and land coverages
Determine how frequently new CIS coverages will be created after core watershed planning
coverages are completed
. Determine how frequently watershed planning databases will be updated
'Develop criteria for integrating and relating data (based on the above findings)
Develop options that meet the criteria (adequate, good, very good) as well as the strengths,
weaknesses, and cost of each option
Get feedback on which option is preferred and fundable (continued)
Exhibit 8. Recommended Milestones and Guiding Policies
-------
No. 5
Designing an Information Management System for Watersheds
(continued from previous page)
Step 5. Evaluate Hardware and Software Configurations
Identify existing plans for reconfiguring hardware and software that may impact information
system design
Determine the priority and sequence of basin planning hardware and software applications
Identify existing hardware and software and how they can best be incorporated into the water-
shed planning information system
Evaluate the compatibility of operating systems (for example, transferring data between the PC
DOS and workstation UNIX environment)
Determine need for exchanging and accessing data (including network speed)
Develop criteria for configuring the hardware/software/network
- Develop optiohs that meet the criteria {adequate, good, very good) as well as the strengths,
weaknesses, and cost of each option
Get appropriate feedback on which option is preferred and fundable
Step 6. Evaluate Organizational Design, Staffing, and Support Issues
Based on preferred hardware/software and database management models, outline information
management responsibilities
Identify staffing needs (including hiring, reassigning, and training staff)
- Outline staffing options (adequate, good, very good), as well as the strengths, weaknesses,
and cost of each option
Get appropriate feedback on which option is preferred and fundable
Step 7. Develop a Short- and Long-Range Implementation Plan
Based on Steps 2-6, develop a multi-phased, 5-year plan, including
Staffing (including training)
- Hardware
- Software
Application development
- Data development, conversion, and integration
Network/communication
Include realistic funding for each component
i1BMHMMMMBM^MMMBMMMMMMMaMH
Exhibit 8. Recommended Milestones and Guiding Policies (continued)
16
-------
Information Transfer Series
Watershed Academy ' ______
application development
data conversion
network/communication
Experience from other states regarding the implementation plan shows that, to be
successful it is important to remember the dynamic nature of an .mplementat.on plan
5 e be flexible to take advantage of emerging grant opportunities or new technolo-
gies and address .short-term needs) at the same time that you work stead.ly^toward
the long-term vision. Exhibit 8 suggests milestones and gu.d.ng pol.c.es for each step
in the process.
17
-------
-------
Watershed Academy ' Information Transfer Series
Appendix A
Definitions for Six Basic Computer System Components
Before comparing GIS systems and hardware/software configurations, it is important
to understand some frequently used terms. Definitions are divided into the six basic
system components: .
Small Bits and Bytes
Processing Units
Input Devices
Output Devices
Mass Storage Devices
Communication Network Lines
Note: Some price ranges listed below may vary by region and will change over time.
Prices are listed for conceptual perspective and should not be used for planning purposes.
Small Bits and Bytes ,
A bit is the smallest unit of data stored in a digital computer and represents the state
of a binary switch {e.g. on or off, 1 or O). A byte is the basic unit of data storage in
computers and consists of eight contiguous bits. Other common data units are a kilobyte
(KB) which is 1024 bytes, a megabyte (MB) which is 1024 kilobytes, and a gigabyte (GB)
which is 1024 megabytes.
Processing Units
Microcomputers: commonly referred to as personal computers or PCS, microcom-
puters normally support one user with a limited number of peripheral devices (e.g.,
printer, modem, etc.). They can also fill the role of a server (an administrative
machine) on a network of microcomputers. A computer's memory (storage space)
is divided into two components: random access memory (RAM) and mass stor-
age. Currently, the most popular PC operating systems are Windows, MacOS, and
OS/2. Baseline PC systems that are currently available generally contain 16 MB of
RAM and approximately 1 GB of mass storage. Such a system typically costs
between $1,500 and $3,500.
19
-------
5 _ Designing an Information Management System for Watersheds
. Workstations: usually support one user plus peripheral devices, are more powerful
than microcomputers, and support multitasking (ability to perform more than one
operation at a time). They are often used as network servers, graphics worksta-
rnsTf or computational intensive tasks. The most popular operating .system .for
workstations is UNIX, although OS/2 and Windows can also be used. Worksta-
tions are available in many system configurations, which can range in price from
$10,000 to $25,000.
Minicomputers: support multiple users (often up to 1 00) working on remote termi-
nate anTaTused primarily as servers in a network of minicomputers, works a-
tions and microcomputers. UNIX and VMS are the most common minicomputer
operating sterns. Minicomputer prices vary depending on their configuration and
manufacturer.
Input Devices
Digitizing Tablets: convert analog data (e.g., maps) into digital information capable
ofS&U into CIS and mapping software packages W**^
tor (line and point) features by tracing analog features (eg., roads, land parcels,
and watershed boundaries) with a digital cursor and record.no *« cootrd'^'°nc»-
tions (i e digitized X and Y points) in a software package. D.gitiz.ng tablets ange
In ?ze and cost from small, desktop models (12" x 12") of about $200 to large,
tabletop models (30" x 40") of over $3,000.
. Optical Scanners: convert analog data into electronic images for input into map-
ping or image manipulation. Scanning uses a photoelectric process to ecord I data
fnto a digital image consisting of unrelated pixels (i.e., picture elements). Scanning
offers a more rapid means to convert analog data into digital data, however fea-
ture identities are lost. Flatbed scanners, the most common type, scan up to legal
size (8.5" x 1 4") and cost from $400 to $ 1 ,000.
Output Devices
Prices for output devices are based on the resolution, color, and size of the > ou Jut
generated, and range from $300 for desktop ink jet printers to $50,000 for E-s,ze (36 x
48") capable, electrostatic plotters.
. InkJet Printers: useful for low-cost, low-resolution color output. Limited to letter
and legal size output.
Laser Printers: useful for low-cost, high resolution, poend
laser printers are available at significantly increased prices. Um.ted to letter and
legal size output.
Pen Plotters: useful for generating extremely high resolul for.,
Hnework and simple features and patterns on paper sizes A (8.5 x 1 Tl ) to . E (3b
x 48"). Limitations include a finite palette (limited color set and patterns) and
susceptibility to media errors (e.g., pen artifacts in Imework and patterns, paper
tears, pens running out of ink).
20
-------
Information Transfer Series
Ink Jet Plotters: generate high resolution, color output on paper sizes A to E.
Electrostatic Plotters: generate high resolution, color output on paper sizes A to E
(or larger). Limited by high cost of hardware.
Mass Storage Devices
Disk Drives: peripheral device storage connected to the central processing unit,
which stores data and applications (software) for direct access. Disk drives fall
into three categories: hard disks (including disk arrays), floppy disks, and remov-
able media (e.g., optical disks, Syquest cartridges, and Iomega Jaz drive cartridges).
Removable storage capacity ranges from 1 .4 MB to 1 GB and fixed sto rage capaci-
ties range from 400 MB to 12 GB. Disk drives range in ^^Tnn^'0^ PR
$50 for a floppy disk drive, $200 for a 1 .2 GB hard disk, to $1 ,500 for a 2.6 GB
removable optical hard drive. Removable media range in cost from about $ 1 for a
floppy disk (1 .4 MB), $40 for a Syquest cartridge (230 MB), and $200 for an
optical rewritable media cartridge (2.6 GB).
Tape Drives- high speed data archiving devices which store data in compressed
format on magnetic tapes. Common magnetic tape sizes are 3mm, 4mm, 8mm,
and 9mm, with capacities ranging from 40 MB to 1 0 GB. Tape drives range ,n cost
from about $200 for low capacity tape drives to about a $1000 for high capacity
tape drives. Tape cost ranges from $5 for low capacity tapes to $20 for high
capacity tapes.
CD-ROM Drives: CD-ROM technology has storage capacities from 660 MB to 2.3
GB CD-ROM drives cost from less than $ 1 00 (for slow speed) to $400 (for high
speed). Recordable CD-ROM drives are also available at an increased price (about
$ 1 ,000) to write to CD-ROMs.
Communication Network Lines:
. Ethernet: communications standard used in many local ^^t^'j^'rANs)
"undreds of protocols may be used with Ethernet, such as TCP/IP, IPX (Novel I),
Appletalk/and DECNet. Ethernet has a throughput of approximately 10 megabits
per second (Mbps), and may use 1 0Base-5 (thick coax), 1 0Base-2 (thin coax) or
1 0Base-T (twisted pair) cable. Data transfer between sites may be accomplished
using ISDN, frame relay, or leased lines.
. Integrated Services Digital Network (ISDN): uses existing copper telephone lines
to provide simultaneous voice and/or data communications. The Basic Rate Inter-
face .(BRI), the most common configuration capable of supporting up ,to three
,
simultaneous voice or data "conversations", is usually not ,
tinuous network communications. Transmission rates on ISDN vary frorr ,64 to
1 28 Kilobits per second (Kbps). To incorporate ISDN into a LAN an ISDN outer is
required at a" cost of approximately $1,500 to $2£00. In addition telephone
charges for a BRI configuration range from $25 to $70 per month, plus $ 01 to
$ 04 per minute of use. Flat rates are available in some areas. Fmally, Internet
access using ISDN generally costs about $200 per month.
27
-------
Designing an Information Management System for Watersheds
Frame Relay: a cost-effective alternative to traditional leased-line connections. It
uses technology based on the efficient switching of data packets across the net-
work It is less costly than the equivalent service provided by leased lines. Frame
relay is available in capacities ranging from 56 Kbps to 1,5 Mbps. At higher capaci-
ties frame relay will support dozens of simultaneous users. Costs for using frame
relay include telephone charges from $175 to $475 per month and an on-site
router for about $2,000. Internet access charges for frame relay range from $200
to $800 per month, depending on the transmission rate.
Leased Lines: provide continuous point-to-point network access. They consist of
dedicated trunk lines between user locations and telephone company central of-
fices Leased lines are generally available in capacities ranging from 56 Kbps to 10
Mbps. Telephone charges vary according to the distance from the central tele-
phone company office and the capacity of the line. They typically range from $250
to over $1,000 per month. On-site equipment (e.g., a router) ranges from $2,000
to $4,000. internet access fees range from $500 to $1,200 per month.
In determining the appropriate communications link, the watershed partners need to
consider the extent to which real-time data access is needed; the degree of data central-
ization; existing hardware, software, and network configurations;^ available resources
Graphic-intensive applications such as GIS typically require a T1 line for fast, on-line
access.
22
------- |