MBTC  2031 – GIS Based BMP Planning Tool for Stormwater Quality Management



Dr. Steve Burian                                          Dr. Findlay Edwards

4190 Bell                                                     4190 Bell

University of Arkansas                                 University of Arkansas

Fayetteville, AR 72701                                Fayetteville, AR 72701



Arkansas Highway and Transportation Department



The objective of the proposed research project is to develop a tool within a geographic information system (GIS) environment to improve stormwater quality management planning for transportation systems.



Stormwater quality management has been an integral part of planning, design, construction, and maintenance of transportation systems since the 1970s.  But only in the 1990s have local, state, and national stormwater regulatory elements become widely applicable to transportation systems.  Specifically, during the past decade the National Storm Water Program has focused on stormwater dischargers with two phases of federal regulations.  The first phase was promulgated on November 16, 1990 and requires selected industries, construction sites disturbing greater than five acres, and municipal separate storm sewer systems (MS4s) serving more than 100,000 people to obtain National Pollutant Discharge Elimination System (NPDES) permit coverage for stormwater discharges.  Permit coverage requires the development of a stormwater management plan and monitoring in some cases.  The second phase is scheduled to go into effect on March 10, 2003, at which time several additional industries, construction sites disturbing at least one acre, and small regulated MS4s will also be required to obtain NPDES stormwater permits.  Current and impending stormwater regulations are applicable to transportation systems during planning, construction, management, and operation.


Given the increased regulations, potential environmental impacts, and possible litigation it is imperative that stormwater runoff from transportation systems be managed proactively in an effective manner.  Moreover, because of the massive potential costs associated with BMP implementation, cost-effective stormwater quality management plans must be developed.  The objective of the proposed research project is to develop a tool within a geographic information system (GIS) environment to improve stormwater quality management planning for transportation systems.  The GIS tool will contain recently collected best management practice (BMP) information on applicability, pollutant removal performance, cost, operation, and maintenance.  The GIS tool will use an expert system procedure to aid the transportation planner and design engineer in the process of BMP selection, performance evaluation, and cost assessment.  The GIS tool will be applicable for existing system stormwater retrofits and new construction of a wide array of transportation systems (e.g., highways, airports, railroad yards) in rural and urban settings.  Overall, the proposed project will take advantage of the widespread use of GIS and the growing body of BMP cost and performance data in creating a convenient tool for design professionals.



Task 1: BMP data collection

The first research task will be to collect and organize information related to stormwater BMPs for transportation systems.  We will draw from several existing sources of BMP pollutant removal performance information including the National Stormwater Best Management Practices (BMP) Database (http://www.bmpdatabase.org/) and the National Pollutant Removal Performance Database [Winer, 2000].  We will also incorporate recent research results published in summary reports [e.g., Sullivan and Borst, 2001] and obtained in unpublished form from researchers.  Cost information will be obtained from numerous national surveys [e.g., Sample et al., 2001], municipal and state authorities, and product vendors.  We will also contact researchers active in evaluating stormwater BMPs, and incorporate independent testing results for proprietary devices.  The following is a preliminary list of BMP types that will be included in the GIS tool:


§         Infiltration/exfiltration practices

§         Filtering practices

§         Vegetated swales

§         Vegetated buffers

§         Detention/retention ponds (dry and wet)

§         Wetlands

§         Baffle boxes

§         Swirl concentrators


Additional BMPs will be added to the above list as the project progresses.  The following is a preliminary list of BMP characteristics that will be included in the GIS tool:


Screening Factors


§         Land area required

§         Soil type needed


§         Meet regulations

§         Social-political acceptance


§         Peak discharge control

§         Runoff volume control

§         Low flow requirements

§         Distance above groundwater

§         Groundwater recharge

§         Streambank erosion control


§         Wildlife habitat creation

§         Aesthetics

§         Negative impacts (e.g., thermal enhancement, vectors)


Evaluation Factors

§         Pollutant type and removal percent

§         Cost

§         Operation and maintenance requirements


We will determine national information and then incorporate local information for the Northwest Arkansas region as a separate database.  Other users can then use the Northwest Arkansas database as a template to develop their own site-specific information database to use with the GIS tool.


Task 2: GIS tool development

The second research task will be to develop the GIS-based planning tool.  We will use the ESRI ArcView 8.x GIS because it is currently one of the most commonly used desktop GIS and mapping software packages.  The planning tool will be coded in Visual Basic for Applications (VBA), the current standard scripting language in the ArcView GIS 8.x product line.


The first step in the development of the tool will be to select the processing algorithms for terrain analysis, hydrology, and runoff quality.


Terrain analysis is required to process the topography data to determine flow direction, identify sinks, and delineate watersheds [Wilson and Gallant, 2000].  The user will have three options for determining drainage patterns on the site:


1.      Create drainage catchment coverage using outside information and tools

2.      Implement the “D8” flow routing approach

3.      Implement a multiple flow path drainage pattern analysis algorithm


The D8 flow routing approach [Costa-Cabral and Burges, 1994] is the most common automated drainage analysis algorithm [Tribe, 1992].  However, its behavior has been widely recognized as poor in some situations, because it can yield slope lines parallel to grid lines or diagonals rather than parallel to the actual slope direction, and it is unable to model divergent flow appropriately [e.g., Costa-Cabral and Burges, 1994; Fairfield and Leymarie, 1991].  But, it will still be included because of its proven ability to provide adequate catchment delineation in most cases, especially for developed catchments.  The multiple flow path drainage analysis will include treatment of flat areas and depressions and permit multiple flow paths to be followed [Martz and Garbrecht, 2000; Rieger, 2000].  The multiple flow path algorithm will be the recommended method.


The hydrology will be based on a distributed SCS (Soil Conservation Service, now Natural Resources Conservation Service) Curve Number hydrology [SCS, 1986].  A spatially distributed approach will be used where each drainage catchment can be subdivided into smaller contributing areas that can have their own unique curve numbers [Moglen and Beighley, 2002; Grove et al., 1998].  Curve numbers and lag times (times of concentration) will be assigned for each drainage catchment by the user and the tool will compute the long-term runoff volume contributions to each drainage point.


The user will have the option of entering in the local average annual rainfall amount or else selecting a state and county combination and allowing the GIS tool to access a database containing a national summary of average annual rainfall amounts.  The hydrologic calculations will begin by computing the average annual runoff volume from each land use type in each drainage catchment included in the GIS database.  The total runoff volume from the catchment will be found by summing the contributions from each subcatchment in the catchment.


The water quality will be simulated using assigned event mean concentrations (EMCs) for each land use/land cover type in the drainage catchment.  The EMC is a measurement of the flow-weighted concentration of a particular constituent (e.g., suspended solids, nitrate) [EPA, 1983; Schueler, 1987].  Runoff quality simulation with the EMC procedures requires the input of appropriate flow-weighted EMCs for each of the pollutants for each of the land classes being simulated. Use of EMCs is appropriate for most long-term studies of stormwater impacts on receiving waters because most larger receiving water bodies respond slowly to storm inflows compared to the rate of change of contaminant concentrations during a storm event [Charbeneau and Barrett, 1998].  Under these circumstances, the event load or EMC is more important than the exact contaminant concentrations during the storm event.  This is especially true for nutrients and solids, but not for contaminants where acute toxicity is important (e.g., dissolved metals, toxicants).  Using an average EMC from several monitored runoff events is appropriate for long-term simulations even though EMCs typically vary from storm to storm.  The total load estimated with EMCs for a long-term simulation is usually as accurate as that obtained from more complex models, e.g., buildup-washoff.


To predict the average annual pollutant load, the product of the average EMC and the average annual runoff volume for each land use type will be computed.  The average annual pollutant loads for each land use type will be summed to find the average annual pollutant load from each drainage catchment:



where Lj is the average annual pollutant load for catchment j [ kg], EMCi is the average event mean concentration for land use type i [mg/l], Voli is the average annual runoff volume [ft3], 28.312x10-5 is a unit conversion, and N is the number of land use types in catchment j.


Following the computation of the average annual load, the GIS tool expert system will be initiated and the user will be guided through four steps: (1) BMP screening, (2) BMP location, (3) BMP pollutant removal performance evaluation, and (4) cost estimating.  The GIS tool expert system will be programmed with accepted BMP screening rules [e.g., ASCE, 2001; WEF and ASCE, 1998; Schueler, 1987].  The user will select the BMPs to place in each catchment.  The tool will then compute the modified average annual runoff volume and pollutant load using the BMP information control data contained in the database developed during research task #1.


The simplistic approach taken to predict runoff volume and pollutant load is appropriate for the planning tool we are developing because we are interested in comparing the relative runoff volume and pollutant load reductions produced by alternative BMP plans.


Task 3: GIS Tool Evaluation

The final research task will be the evaluation, assessment, and revision of the GIS tool.  An operational version of the tool will be distributed to a selected set of undergraduate students, graduate students, academics, and design professionals with varying levels of experience in stormwater quality management and transportation system planning.  Using a cross section of experience levels will enable the convenience and ease-of-use of the tool to be assessed.  The evaluators with higher levels of experience will provide assessment of the technical quality and comprehensiveness of the tool.  The evaluators’ recommendations and additional desires will be incorporated into the final version of the tool before final release.




Start Date: July 1, 2002

End Date: June 30, 2004




Year 1: Anticipated Accomplishments

§         produce a report summarizing the performance, cost, advantages, disadvantages, and intangibles pertinent to a suite of stormwater BMPs applicable to transportation systems

§         develop terrain processing, hydrology, and runoff quality algorithms to use in the GIS tool to predict the runoff volume, pollutant load, and BMP performance and cost for a particular drainage catchment and chosen BMP plan

§         initiate the coding process

§         present conceptual ideas of GIS planning tool at local and national conferences to obtain constructive feedback

§         create web site describing GIS tool development project; update web site periodically during the remainder of the project


Year 2: Anticipated Accomplishments

§         finish coding of GIS tool

§         internal testing of GIS tool by graduate students funded by project and in graduate stormwater quality management courses

§         limited distribution of tool to a selected set of outside evaluators

§         modification of tool based on evaluator feedback

§         final release of version 1.0; distribute via web site and direct contacts

§         prepare articles and present papers at local and national conference





MBTC Grant


$ 69,343

Total Budget

$ 198,735




The project will support the PI and Co-PI with academic year and summer salary.  Two graduate students will also be supported during the duration of the project.  The graduate students will receive the standard monthly stipend of $1200.00/month and tuition will be paid.  Travel funds will be used to defray the cots of travel for the project team to attend regional and national conferences to disseminate project results.  Equipment funds will be used to purchase a desktop computer system and one ArcView 8.x license.  A computer system dedicated to the project is necessary for successful project management and completion.  The fringe benefits are assigned rates for personnel employed on research projects.  The indirect cost rate (42.5%) is the assigned value by the University of Arkansas.

The required 50% cost-share funds will be obtained from the Arkansas Highway and Transportation Department (AHTD) and the University of Arkansas Department of Civil Engineering.



The proposed project will capitalize on the experience gained by the project team developing GIS analysis tools for other applications (see discussion in Qualifications of Research Team).  In addition, the project activities will strengthen the GIS development abilities of the project team and aid in the development of new projects.


Members of the project team are currently involved with several stormwater BMP research projects (see discussion in Qualifications of Research Team).  The development of a BMP planning tool during the proposed project will use results from the other studies of BMP performance.



Results and products from the proposed research project will be disseminated through paper presentations at professional meetings at the local, state, and national levels.  Articles will be written for research journals and trade magazines describing the development of the GIS tool and the finished product.  Significant contributions to the body of stormwater quality management will be made with the publication of the summary report of BMP pollutant removal performance and cost information related to transportation systems and the creation of a convenient GIS-based planning tool.

The GIS tool will be obtainable through a web site.  The web site will be created during the first few months of the project and periodically updated as the project progresses.  Upon completion of the GIS tool the web site will be updated in order to permit downloading of the tool script and documentation.

The stormwater and transportation community will be made aware of the availability of the GIS tool through email announcements to various listserves.



If effective environmental controls are not included in the planning and design of transportation systems, the construction and operation can be detrimental to nearby natural resources.  The product of the proposed research project will benefit society by improving the planning and design of stormwater quality BMP plans.  Improved BMP plans will help to protect existing water quality and improve degraded water resources, which has direct economic and quality of life benefits for society.

The stormwater quality management planning and design community will also experience benefits from the results of the proposed project.  The GIS tool will aid in the planning and evaluation of stormwater BMP plans and take advantage of the commonly used ArcView 8.x GIS.  Time and effort for stormwater planning will be reduced and improved designs will be produced.

The transportation community will also benefit from the results of the proposed project.  The GIS tool will improve the ability of planning and design professionals to incorporate stormwater quality management practices into transportation system designs.  The tool will provide the user with a convenient method to screen and select BMPs, and evaluate the performance and cost of alternative BMP plans.

Graduate students at the University of Arkansas will benefit from the proposed project.  Two graduate students will be supported on assistantship to work on the project.  They will learn advanced concepts of stormwater quality management and become versed in advanced GIS programming.   The combination of water resources engineering knowledge and GIS skills will make the graduate students very marketable upon graduation.  Graduate students enrolled in the stormwater quality management courses offered by the project investigators will also benefit from the knowledge they will gain during the courses.