View expanded Pesticide Monitoring MapGROUND WATER PROTECTION
STATE OF INDIANA DRAFT GENERIC
PESTICIDE MANAGEMENT PLAN
Appendix F
A Groundwater Vulnerability to Pesticides Map for Indiana
ABSTRACT:
A technique for developing groundwater vulnerability to pesticides maps for state-sized areas using, geographic information system (GIS) databases and other databases commonly available for such areas will be used to create a groundwater vulnerability map for Indiana. The technique will improve upon existing techniques that estimate groundwater vulnerability as a result of hydrogeologic factors. Available groundwater quality information will be used to statistically evaluate the resulting vulnerabailitly maps.
JUSTIFICATION:
The presence and potential presence of pesticides in groundwater is a serious problem in many locations. The EPA National Survey of Pesticides in Drinking Water Wells, started in 1988, tested a total of 1347 samples fr6m 564 community and 783 rural drinking water wells for nitrates, pesticides, and pesticide breakdown products (USEPA, 1990). The wells, selected statistically, are representative of more than 10.5 million rural wells and more than 94,600 wells in approximately 38,300 community water systems. The survey results indicated approximately 0.8% of the wells contained pesticide residues above levels considered the Maximum Contaminant Level (MCL). However, a significant number of the wells tested contained pesticide residues less than the MCL. In addition, approximately 1.2% of the community wells tested and 2.4% of the rural wells tested contained nitrate levels above the Maximum Contaminant Level (MCL) for nitrate. The nitrate MCL is 10 parts per million. Many other wells tested contained nitrate levels near the MCL. Over 52% of the community water systems and 57% of the rural domestic wells tested contained nitrates (USEPA, 1992).
When viewing the maps, use your "BACK" button to return.
View expanded Nitrate Monitoring Map
View expanded Pesticide Monitoring Map
Statewide maps showing areas that are vulnerable to groundwater contamination from pesticides have many potential uses. Such maps are needed to assist with the implementation of pesticide management plans to prevent degradation of groundwater quality. Vulnerability maps would be useful for implementation of groundwater quality monitoring programs. These maps would be useful for examining existing and potential policies for groundwater protection including zoning and other programs that influence new development. Education programs on groundwater protection could also be tailored for areas based on such maps.
In most states, the only statewide groundwater vulnerability maps that are available are based on the DPAISTIC maps that were produced from 1:2,000,000 data (Aller et al, 1985 and Aller et al., 1987). The USEPA (1992) found that these maps did not correlate well with the water quality analysis performed for tjae National Survey of Pesticides in Drinking Water Wells. More detailed and more accurate maps are needed by states to implement programs such as those described above.
OBJECTIVE:
The objective of this project is to develop and statistically test a statewide groundwater-vulnerability to-pesticides (GVP) map for Indiana More specifically, the objectives are to:
a. Develop a technique that utilizes commonly available GIS (geographic information system) databases for large areas (states) and commonly available water quality databases to create GIS-based groundwater-vulnerability-to-pesticides maps for Indiana.
b. Develop a GIS-based, groundwater-vulnerability-to-pesticides map for the state of Indiana using this technique.
c. Test the groundwater-vulnerability-to-pesticides map developed for Indiana using well water samples that have been analyzed for pesticides.
PROCEDURES:
A technique for developing groundwater-vulnerability-to-pesticides maps for state-sized areas will be developed that uses geographic information system (GIS) and other databases commonly available for such areas. The technique will improve upon existing techniques that estimate groundwater vulnerability as a result of hydrogeologic factors. The improved map will be compared for accuracy against results from previous methods. The technique will be applied to develop a groundwater vulnerability to pesticides map for the state of Indiana. Groundwater quality information will be used to statistically evaluate the resulting vulnerability maps. The procedures are explained in more detail below.
The technique proposed for developing maps of groundwater-vulnerability-to-pesticides will involve several steps. A description of the proposed technique follows. An initial set of vulnerability maps will be developed based on hydrogeologic factors as considered by DRASTIC (Aller et al., 1985 and Aller et al., 1987) and SEEPAGE (Carpenter, 1992). DRASTIC is an empirical model developed by the National Water Well Association in conjunction with the EPA for evaluating groundwater contamination potential on a regional basis. SEEPAGE (System for Early Evaluation of the Pollution Potential of Agricultural Groundwater Enviroments) estimates contamination potential of groundwater using hydrologic factors in a manner similar to DRASTIC. The DRASTIC groundwater vulnerability maps that were prepared for the U.S. were based on 1:2,000,000 scale map data. In this project, DRASTIC and SEEPAGE will be applied with 1:250,000 scale on more detailed map data.
The factors considered in DRASTIC are:
Depth to water
Recharge
Aquifer media
Soil media
Topography
Impact of the vadose zone media
Conductivity (hydraulic) of the aquifer
SEEPAGE considers the following factors:
Soil slope
Depth to water table
Vadose zone material
Aquifer material
Soil depth
Attenuation potential
The SEEPAGE attenuation potential factor further considers the following factors:
Soil surface texture
Subsoil texture
Surface layer pH
Organic matter content of surface
Soil drainage class
Soil permeability
The results of DRASTIC and SEEPAGE will provide initial vulnerability maps based on hydrogeologic factors. Pesticide use maps will be generated using 1:250,000 scale land use/land cover data from the USGS and county pesticide use data. Land use data are available for the entire country at this scale in GIS formats. Crop types as a percentage of crop land within counties will be obtained from state agricultural statistics or the U.S. Agricultural Census database. Pesticide use by county will be obtained from state agencies that collect this data (within Indiana, this data is collected by the Office of the Indiana State Chemist) and the U.S. Agricultural Census database. Using this information, estimates of spatial pesticide application for the state will be obtained. The pesticide distribution map will be generated within a GIS. The initial vulnerability maps and the pesticide use map will be combined using GIS techniques to develop a new map showing both vulnerability and pesticide use.
When viewing the maps, use your "BACK" button to return.
View expanded DRASTIC Pesticide Vulnerability Map
In the next step, the maps showing vulnerability and pesticide use will be compared with data from a national well location and well pesticide concentration database (EarthInfo Inc, 1992) and to state (or regional) databases of similar data. Within Indiana, the EarthInfo database contains approximately 3,500 data points, the Farm Bureau Indiana Cooperative Well Water Testing Program (1992) database contains approximately 8,000 data points, and the USGS pesticides in wells database contain an additional 875 data points. Fifty percent of the data points in these databases will be used for the comparison. The remaining data will be reserved for statistically analyzing the final vulnerability maps. The comparison of the hydrogeologic factors and pesticide use maps with the pesticide concentrations in groundwater databases will be used to identify large areas (greater than 25 percent of a county) in which the maps and data disagree.
For the areas of disagreement, the pesticide concentration data will be examined to determine possible causes for disagreement. For some data points, site descriptions and information indicate the probable cause of high pesticide concentrations. Using this information, some areas may be eliminated from further consideration if causes of pesticides are not likely a result of hydrogeologic factors and/or pesticide applications. For example, problems that are caused as a result of well condition will be eliminated.
Once such areas are eliminated, GLEAMS, RUSTIC, or PRZM will be applied to the remaining areas of disagreement to examine the simulated movement of pesticides into the groundwater. Because of the time and expense that would be required to run these models for all areas within a state, only areas of significant disagreement will be examined. However, the technique described here would allow additional areas to be simulated using these process-oriented models in the future. For example, a county might use the models to improve groundwater vulnerability maps for a portion or all of the county at some future date. GLEAMS (Groundwater Loading Effects of Agricultural Management Systems) (Leonard et al, 1987), RUSTIC (Risk of Unsaturated/Saturated Transport and Transformation of Chemical Concentration model) (Donigian, 1990) and PRZM (pesticide Root Zone Model) (Carsel et al., 1984 and 1985) simulate both hydrology and chemical transport and are capable of providing more detailed assessments of contamination potential than DRASTIC and SEEPAGE.
The data necessary to run these models will come from the GIS data layers, water quality database site descriptions, paper maps, and visits to the areas of disagreement. If the models indicate a different vulnerability, the site GVP maps will be adjusted. The vulnerability maps will also be adjusted to a greater vulnerability if groundwater quality data indicate consistent pesticide detection within a region and hydrogeologic vulnerability estimates are low. Model results and data from the groundwater pesticide database will be used to modify the GVP map to produce final vulnerability maps.
The vulnerability maps produced will be statistically analyzed using the groundwater pesticide data that was not used for development of the final vulnerability map (50% of the pesticide groundwater data points). These data come from the EarthInfo Inc. (1992) CD ROM, the Farm Bureau Indiana Cooperative Well Water Testing Program (1992) database, and USGS databases of similar data. The model-generated maps that will be analyzed include:
DRASTIC
SEEPAGE
DRASTIC plus pesticide use
SEEPAGE plus pesticide use
DRASTIC plus pesticide use and model results
SEEPAGE plus pesticide use and model results
DRASTIC plus pesticide use and groundwater pesticide data
SEEPAGE plus pesticide use and groundwater pesticide data
DRASTIC plus pesticide use, model results, and groundwater pesticide data
SEEPAGE plus pesticide use, model results, and groundwater pesticide data
It is anticipated that the DRASTIC and SEEPAGE vulnerability maps that consider pesticide use, groundwater and chemical movement simulations, and groundwater pesticide data will be significantly better than the vulnerability maps that consider less data.
The proposed technique for developing statewide groundwater vulnerability-to-pesticide maps (described above) will be applied to the entire state of Indiana. The spatial data required to generate the initial vulnerability maps using DRASTIC and SEEPAGE are available in GIS and paper map formats. The DRASTIC map layers for the soil media, topography, and impact of vadose zone media are currently available for the state of Indiana in the GRASS GIS format in the Agricultural Engineering Department at Purdue University. The SEEPAGE map layers for the slope, vadose zone material, soil depth, and attenuation potential factors are currently available for the state of Indiana in the GRASS GIS format in the Agricultural Engineering Department at Purdue University. The soil media map was digitized from the 1:200,000 soil association maps for the state of Indiana. In addition, the SCS STATSGO GIS soil data are available for Indiana. The topography layer was derived from the DMA/USGS digital elevation model data that came from the DMA 1:250,000 elevation maps. The impact of vadose zone media layer and other soil related data layers were derived from the soil association map and the Soils 5 database.
The remaining data layers needed to run DRASTIC and SEEPAGE will be digitized from paper maps or developed from other spatial databases to provide GIS layers. Since depth to groundwater maps do not exist for the entire state, the depth to water layer will be developed from regional maps and point data for depth to water. Databases of depth to water at points are available from the USGS and EarthInfo (1992). The aquifer recharge map will be digitized from existing paper maps. The USGS "Atlas of Indiana Aquifers" will be used to develop the aquifer media map layer. The Hydraulic Conductivity of aquifers map layer will be developed from hydraulic conductivity point data and the aquifer type map layer. The GIS tool can assist with the process of generalizing point and limited data to remaining areas for the state. Similar databases and maps exist for other states and could be readily used to develop the data required for the groundwater vulnerability estimation technique described here.
Once the required DRASTIC and SEEPAGE layers are available in the GRASS GIS format, they will be weighted according to DRASTIC and SEEPAGE procedures and the weighted layers will be summed within the GIS to, provide the initial groundwater vulnerability map layers. Areas with a high vulnerability rating and high pesticide application rates would be the most critical areas.
Using the DRASTIC and SEEPAGE results along with pesticide use by county and a statewide land use/land cover GIS layer (as described above), a new set of maps will be produced showing vulnerability (based on hydrogeologic factors) and pesticide use.
The resulting maps from the above step will be analyzed using 50% of the groundwater pesticide concentration data available for Indiana. These groundwater pesticide concentration data will be randomly selected from the EarthInfo Inc. (1992) CD ROM, the Farm Bureau Indiana Cooperative Well Water Testing Program (1992). database, and from USGS data sets collected within the state. All of the data points in these databases have the location coordinates from which the sample was collected. The groundwater pesticide concentration data will be used to determine areas in which the vulnerability map disagrees with these data.
The areas of significant disagreement will be identified and further analysis of these areas performed as described above. The pesticide data for areas of disagreement will be examined to determine if a possible reason for contamination was included in the database. The most appropriate model (RUSTIC, G S, or MM will be run for areas in which disagreements result. Using the model results and pesticide in groundwater data in addition to the hydrologic factors and pesticide use, modified vulnerability maps will be prepared as described previously.
The series of Indiana groundwater vulnerability maps that are produced (see above for a list of maps) will be statistically analyzed using the remaining groundwater pesticide database points (50% of the total points) from the EarthInfo Inc. (1992) CD ROM, the Farm Bureau Indiana Cooperative Well Water Testing Program (1992) database, and from USGS data sets collected within the state. These data along with coordinates of the points from which the water samples were collected reside in a database, allowing quick analyses of the vulnerability maps.
The best Indiana groundwater vulnerability to pesticide map will be provided to the Office of the Indiana State Chemist for use in implementation of a state pesticide management plan. The map will be provided in both paper and GIS formats. The GIS data layers will also be provided to allow future modifications if desired and to allow development of vulnerability maps to other substances. In addition, a report describing the development process, the statistical testing, and results of the statistical analyses will be provided.
LITERATURE REVIEW:
DRASTIC and SEEPAGE
Aller et al. (1987) developed the DRASTIC tool to estimate groundwater vulnerability for large areas based on hydrogeologic factors. DRASTIC is an empirical model developed by the National Water Well Association in conjunction with the EPA for evaluating groundwater contamination potential on a regional basis. The factors considered in DRASTIC are:
· Depth to water
· Recharge
· Aquifer media
· Soil media
· Topography
· Impact of the vadose zone media
· Conductivity (hydraulic) of the aquifer
Areas to be rated by DRASTIC are subdivided into areas in which the factors considered are nearly homogeneous. A grid cell subdivision is often used and thus raster GIS are well suited for the application of DRASTIC. Each of the subdivisions is weighted based on each of the above factor values. To develop the DRASTIC score, the weights for each subdivision are summed to obtain the DRASTIC score (Aller et al., 1985). The numerical scores are converted into qualitative risk categories of low, moderate or high.
DRASTIC was applied to the U.S. using maps at a scale of 1:2,000,000 (Aller et al, 1987). The USEPA (1992) analyzed the results of the National Survey of Pesticides in Drinking Water Wells dam and the DRASTIC predictions developed at this scale. County level DRASTIC scores (an aggregated score for each county) and subscores were computed and 90 counties selected for analysis with data from wells sampled in the study. The results showed that DRASTIC performed very poorly for the 90 counties tested. For implementation of pesticide management plans at a state level, a more detailed vulnerability map is needed. The proposed project will use more detailed data for estimating DRASTIC scores and additional data to improve vulnerability estimates.
SEEPAGE (System for Early Evaluation of the Pollution Potential of Agricultural Groundwater Environments) (Carpenter, 1992) is used to evaluate the potential for groundwater contamination from both point and non-point sources considering hydrologic factors. SEEPAGE considers hydrologic factors to locate areas with low, moderate, high, and very high potential for groundwater pollution using GIS data. SEEPAGE considers the following factors:
* Soil slope
* Depth to water table
* Vadose zone material
* Aquifer material
* Soil depth
* Attenuation potential
The attenuation potential factor further considers the following factors:
Soil surface texture
Subsoil texture
Surface layer pH
Organic matter content of surface
Soil drainage class
Soil permeability
These factors are combined using a weighting scheme described in detail by Carpenter (1992). This approach is similar to that used in the DRASTIC model (Aller et al, 1987). For each factor considered, weights are assigned to possible values of the factor. For example, the soil slope factor has the possible values shown in the table below and associated point source and non-point source weights.
SEEPAGE Soil Slope Factor Weights
Percent Slope Point Source Non-Point Source,
0-2 10 30
2-6 9 27
6-9 5 15.
9-12 3 9
>12 1 3
PUMPS
Pickus and Hewitt (1992) developed the Pesticide User Management Planning System (PLWS) to assist resource managers in understanding the effects of pesticides on groundwater. PUMPS uses GIS and the pesticide leaching index model (Leaching Pesticide Index - LPI) developed by Meeks and Dean (1990) to assess groundwater sensitivity to pesticides. The LPI model estimates leaching potential using chemical properties, soil information, crop information, and hydrogeologic data. PLWS is most applicable to small areas because of the detailed data required for operation of the LPI model. PUMPS was demonstrated for a county in Delaware but data for validation were not examined.
Groundwater and Chemical Movement Models
More detailed process oriented models play an important role in simulating the movement of chemicals into the groundwater The GLEAMS (Groundwater Loading Effects of Agricultural Management Systems) model (Leonard et al, 1987) has been shown to be an effective tool la assessing potential pesticide and nutrient leaching below the root zone for agricultural production systems in the Southern United States. GLEAMS also performed well for Indiana conditions in research conducted by Amin et al (1991).
Other process oriented models such as RUSTIC and PRZM have been used to examine groundwater and chemical movement. RUSTIC (Risk of Unsaturated/Saturated Transport and Transformation of Chemical Concentration model) (Donigian, 1990) and PRZM (Pesticide Root Zone Model) (Carsel et al., 1984 and 1985) simulate both hydrology and chemical transport and are capable of providing more detailed assessments of contamination potential than DRASTIC or SEEPAGE. The primary limitation of these models are the detailed data required for simulation.
GIS and Decision Support Applications to Groundwater Protection
Researchers have used GIS-techniques to develop tools to reduce groundwater contamination potential. Zhang et al. (1990) linked a solute transport model with a GIS to assess the potential groundwater impacts of common agricultural chemicals for an Oklahoma watershed. Halliday and Wolfe (1990) implemented a GIS-based decision support system that used DRASTIC and other models to assess the groundwater pollution potential from fertilizers in Texas. Evans and Myers (1990) also implemented DRASTIC in a GIS to evaluate regional groundwater pollution potential. Bleecker et al. (1990) integrated chemical movement simulations with GIS to map groundwater contamination potential. Brickford et al. (1990) used a GIS for county-wide environmental management with emphasis placed on groundwater. Bruner et al. (1990) used a GIS to assist with assessing shallow aquifer vulnerability to groundwater contamination. Focazio (1990) used a GIS to analyze groundwater resources in the Virginia coastal plain. Robillard (1990) linked an expert system with a GIS to develop, a tool for water resources management.
Dr. Engel and his students are currently completing decision support systems for evaluating groundwater quality for a USDA-CSRS project started in 1990 (Decision Support Systems for Evaluating Groundwater Quality Problems) (Embleton and Engel, 1992). The decision support systems consists of expert systems, simulations, and hypermedia modules. Field data is being collected to assist with validation of the decision support systems. The decision support systems will answer the following questions:
Does this location have a groundwater quality problem?
If so, what is the likely source of the contamination?
What corrective measures should be taken?
What is the potential for having a groundwater quality problem at this location?
What measures can be taken to reduce chances or prevent contamination of groundwater at this location?
GIS and Watershed Modeling
Dr. Engel has developed several spatially-based decision support systems that utilize geographic information systems (GIS), expert systems, neural networks, and watershed simulations to estimate runoff, erosion, and surface water chemical movement (Srinivasan and Engel 1991 and Engel et al., 1992). Engel et al (1990) linked ANSWERS (Areal Nonpoint Source Watershed Environmental Response Simulation) to the GRASS (Geographic Resources Analysis Support System) (U.S. Army, 1989) GIS to overcome the time and expense required to collect input data for ANSWERS and to provide a more controlled and explainable set of rules for selecting specific information in the ANSWERS data file.
The integrated system was tested on a small watershed within the Indian Pine Natural Resources Field Station (,Engel et al, 1990). The ANSWERS/GRASS link significantly reduced the time, expense, and expertise required to collect inputs for ANSWERS. Display of the simulated results using the GIS provided an effective means for interpretation In addition, the results could be analyzed with other layers of information that are stored in the GIS.
Arnold et al. (1990) integrated simulations and knowledge systems (expert systems) with a GIS to assist with industrial site suitability analysis with emphasis on potential environmental impacts. They also examined the possibility of using neural networks for site selection. As part of the project, they developed water resources decision support tools that were integrated with a GIS to assist in defining, drainage lengths, slopes, and times of concentration.
Groundwater Vulnerability Mapping Within Indiana
The following describe some of the efforts to define groundwater vulnerability within Indiana. The EPA (1990) developed a groundwater vulnerability map to pesticides for Indiana at the county level using a composite DRASTIC score and county pesticide data use. The USDA built on this approach by considering pesticide data on leachability and also incorporated county pesticide use data (Nielsen and Lee, 1987), The maps produced provide a groundwater contamination potential for an entire county. Additional detail is required for implementation of a State pesticide management plan. The DNR has developed Water Resource Assessments for the St. Joseph River, the Kankakee River, and the Whitewater River basins. These assessments include a relative vulnerability to contamination for each aquifer system. The limitation of the assessments is that they exist for only a very small portion of the state. The IDEM Groundwater Section has developed 1:250,000 scale paper maps for the state which delineate and rank areas vulnerable to groundwater contamination based on geology. Other factors and data should be considered in developing defensible groundwater vulnerability maps.
FACILITIES AND EQUIPMENT
This project will utilize existing computer and GIS systems available in the Agricultural Engineering Department at Purdue University. The computer equipment that will be used consists of SUN SPARC I and SPARC II UNIX workstations and SUN file servers with over 30 gigabytes of mass storage. Three digitizing tablets for digitizing paper maps are available to complete the remaining GIS layers needed for this project The GRASS GIS software (U.S. Army, 1989) is available for all of the SUN workstations within the Agricultural Engineering Department. The ARC/INFO GIS tool is also available on these machines. A GRASS GIS data set has been assembled for the state of Indiana during the past 3 years.
These data will be used in this project to assist with the development of the groundwater vulnerability maps.
The Farm Bureau Indiana Cooperative Well Water Testing Program (1992) database contains approximately 8,000 data points from wells within Indiana that provide information about the concentrations of pesticides in these wells. The USGS has compiled a database of pesticide analyses of groundwater samples collected by government agencies in Indiana, including IDEM IDNR, USEPA, and USGS. At the present time, the database contains 875 samples collected from 519 wells between December, 1985 and April, 1991. Detections of pesticides were made in 6 percent of the samples and 8 percent of the wells.
EarthInfo (1992) has assembled various data sets (largely USGS data) to develop CD ROM based databases of groundwater data for the US. The latitudes and longitudes of each well are also recorded in the database. These data will be used in conjunction with the GIS data to develop state groundwater vulnerability maps and for testing vulnerability maps.
LITERATURE CITATIONS:
1. Aller, L., T. Bennett, J.H. Lehr, and R.J. Petty. 1985. DRASTIC: A standardized system for evaluating groundwater pollution potential using hydrogeologic settings. U.S. EPA, Robert S. Kerr Environmental Research Laboratory, Ada, OK, EPA/600/2-85/0108, 163 pp.
2. Aller, L., T. Bennet, J.H. Lehr, RJ. Petty, and G. Hacket 1987. DRASTIC: A standardized system for evaluating groundwater pollution potential using hydrogeologic settings. EPA-600/2-85/0108, 163pp..
3. Amin Sichani, S., B.A. Engel, E.J. Monke, J.D. EigeL and EJ. Kladivko. 1991. Validating GLEAMS with herbicide and insecticide field data on a Clermont silt loam soil. TRANSACTONS of the AS.AE 34(4):1732-1737.
4. Arnold, J.G., X. Zhuang, B.A. Engel, R. Srinivasan, C.C. Rewerts, and R.S. Muttiah. 1990. Intelligent GIS for natural resource modeling and site selection. In: Remote Sensing and GIS Applications to Non-point Source Planning, United States Environmental Protection Agency, Chicago, IL.
5. . Bleecker, M, J.L. Hutson, and S.W. Waltman. 1990. Mapping groundwater contamination potential using integrated simulation modeling and GIS. In: Proceedings of Application of Geographic Information Systems, Simulation Models, and Knowledge-based Systems for Landuse Management. VPI, Blacksburg, VA. pp. 319-328.
6. Brickford, S.L., M.D. Smolen, L.E. Danielson, and H.A. Devine. 1990. Development of a water quality database and assessment strategy for county-level environmental management. In: Proceedings of Application of Geographic Information Systems, Simulation Models, and Knowledge-based Systems for Landuse Management VPI, Blacks . burg, VA. pp. 277-286.
7. Bruner, B.G., G.W. Wilkerson, and J.C. Nye. 1990. Using geographic information system technology to assess shallow aquifer vulnerability to ground water contamination. In: Proceedings of Application of Geographic Information Systems, Simulation Models, and Knowledge-based Systems for Landuse Management. VPI, Blacksburg, VA- pp. 451-460.
8. Carsel, R.F., L.A. Mulkey, M.N. Lorber, and L.B. Baskin. 1985. The pesticide root zone model (PRZM): A procedure for evaluating leeching threats to groundwater. Ecological Modeling 30:49-69.
9. Carsel, R.F., C.N. Smith, LA Mulkey, J.D. Dean, and P. Jowise. 1984. User's manual for PRZM: Release I. U.S. EPA, Environmental Research Laboratory. Athens, GA. EPA-600-3-84-109.
10 EarthInfo Inc. 1992. USGS Quality of Water. CD ROM Database, Boulder, Colorado.
11. Engel, B.A., J.G. Arnold, R. Srinivasan, Y, Zhuang, C. Rewerts, and R. Muttiah. 1990. Runoff, erosion, and chemical movement simulation using GIS. In: Proceedings of State of Indiana GIS Conference. University GIS Alliance, NRRI, Purdue Univ. p. 162-169.
12. Engel, B.A-, R. Srinivasan, and C. Rewerts. 1992. A Spatial Decision Support System for Modeling and Managing Agricultural Non-Point Source Pollution. In: Integrating Environmental Modeling and GIS, NCGIA, Santa Barbara, CA (In Press).
13. Embleton, K.M. and B.A- Engel 1992. An artificial intelligence based farmstead assessment program for groundwater quality concerns. 4th International Conference on Computers in Extension. pp. 750-755.
14. Evans, B.M. and W.L. Myers. 1990. A GIS-based approach to evaluating regional Groundwater pollution potential with DRASTIC. Journal of Soil and Water Conservation 45(2):242-245.
15. Focazio, NiJ. 1990. Application of a geographic information system to the analysis of groundwater resources in the coastal plain of Virginia. In: Proceedings of Application of Geographic Information Systems, Simulation Models, and Knowledge-based Systems for Landuse Management VPI, Blacksburg, VA. pp. 351-359.
16. Halliday, S.L. and M.L. Wolfe. 1990. Assessing groundwater pollution potential from agricultural chemicals using a GIS. Paper 903023, ASAE, St. Joseph, NH.
17. Indiana Cooperative Water Well Testing Project 1992. Symposium on the Indiana cooperative water well testing project. Indianapolis, IN.
18. Leonard, R.A-, W.G. Knisel, and D.A. StilL 1987. GLEAMS: Groundwater Loading Effects of Agricultural Management Systems. Transactions of ASAE. 30(5):1403-1418.
19. Meeks, YJ. and J.D. Dean. 1990. Evaluating groundwater vulnerability to pesticides. Journal of Water Resources Planning and Management 116(5):693-707.
20. Nielsen, E.G. and L.K. Lee. 1987. The magnitude and costs of groundwater contamination from agricultural chemicals: A national perspective. Resources and Technology Division, Economic Research Service, U.S. Department of Agriculture. Agricultural Economic Report -No. 576.
21. Pickus, J. and M. Hewitt. 1992. Resource at risk: Analyzing sensitivity of groundwater to pesticides. Geo Info Systems 2(10):50-55.
22. Robillard, P.D. 1990. Linking GIS to expert systems for water resources management. In: Proceedings of Application of Geographic Information Systems, Simulation Models, and Knowledge-based Systems for Landuse Management. VPI, Blacksburg, VA. pp. 1- 10.
23. U.S. Army. 1989. GRASS reference manual. USA CERL, Champaign, EL.
24. USEPA. 1990. National Pesticide Survey: Phase I Report. EPA, Washington, DC.
25. USEPA. 1992. Another Look: National Survey of Pesticides in Drinking Water Wells. Phase II Report. EPA,579/09-91-020. EPA, Washington, DC.
26. Zhang, H., D. Nofziger, and C.T. Haan. 1990. Interfacing a root-zone transport model with GIS. Paper 903034, ASAE, St. Joseph, NE.
