Scientific Basis for Identifying Ground Water Protection Areas
How DPR identified areas vulnerable to pesticide movement to groundwater
Pesticides have been found in a variety of climatic, soil, and depth to groundwater conditions. In the 1990s DPR scientists wondered whether there was a way to characterize these various conditions so that pesticide movement to groundwater could be predicted, and thus prevented before contamination occurs. The following sections describe the source of pesticide detection data, and the various statistical techniques, the mathematical model, and depth-to-groundwater data used by DPR to identify vulnerable areas. These vulnerable areas where unmitigated use of certain pesticides is likely to contaminate groundwater were adopted in regulation as "Ground Water Protection Areas."
Source of pesticide detection data: Well inventory database
The Pesticide Contamination Prevention Act (PCPA) requires DPR to sample wells for pesticides with the potential to pollute groundwater. The PCPA further requires all local, county and state agencies to submit all results of well sampling for pesticides to DPR. In addition, DPR encourages other entities, such as the United States Geological Service, to submit results. DPR is also required to maintain a database of these well sampling results. As a result, DPR has a large data set of wells containing pesticide residues from legal agricultural use. DPR used this data set to determine if there was a relationship between detections and local climate, soil, and depth-to-groundwater characteristics. DPR's well inventory database and well sampling strategies are summarized in:
- Summary of Well Water Sampling in California to Detect Pesticide Residues Resulting from Nonpoint-Source Applications, PDF. Troiano, J., D. Weaver, J. Marade, F. Spurlock, M. Pepple, C. Nordmark, D. Bartkowiak. J. Environ. Qual. 30:448-459. 2001. Reprinted with the permission of the American Agronomy Society.
Statistical methods and modeling
Because of the wide range of climatic and soil conditions associated with pesticide contamination of groundwater, an empirical statistical approach was devised to determine whether vulnerable areas could be described. The empirical analysis was conducted in two steps. In the first step, a cluster statistical procedure was used to identify groups of contaminated wells with similar soil properties. Initially, principal component analysis was used to associate sections of land with those clusters. These procedures are described in the following two publications - a DPR report and a journal article.
- Profiling Areas Vulnerable to Ground Water Contamination by Pesticides in California, PDF. Troiano, J., B. Johnson, S. Powell, and S. Schoenig. EH 92-09. 1992.
- Use of Cluster and Principal Component Analyses to Profile Areas in California Where Ground Water Has Been Contaminated by Pesticides. Troiano, J., B.R. Johnson, and S. Powell. Environ. Monitor. Assess. 32: 269-288. 1994.
Later, an alternative procedure (canonical variates analysis) was used to improve characterization of vulnerable areas. The following two publications describe how DPR tested that procedure by sampling wells inside and outside of vulnerable areas, considering depth-to-groundwater as an additional factor. The statistical model developed to identify vulnerable areas is called the CALVUL model.
- Profiling areas of Ground Water Contamination by Pesticides in California: Phase II - Evaluation and Modification of a Statistical Model. Troiano, J., C. Nordmark, T. Barry, and B. Johnson. Environ. Monitor. Assess. 45:301-318. 1997.
- Pesticide Movement to Ground water: Application of Areal Vulnerability Assessments and Well Monitoring to Mitigation Measures. Troiano, J., C. Nordmark, T. Barry, B. Johnson, and F. Spurlock. In Ballatine et al. (ed.) Triazine Herbicides Risk Assessment. ACS Symposium Series 683. p.239-251. 1998.
Application of the CALVUL model
Previously, DPR sampled for pesticides in groundwater based on pesticide use data only. However, in determining where to sample groundwater for norflurazon residues, DPR used the CALVUL model and pesticide use data. High norflurazon detection rates indicated that the CALVUL model is an effective tool to identify areas where contamination is likely. The norflurazon study also indicated that depth-to-groundwater was another important variable in determining vulnerable areas. These results are found in the following report:
- Empirical Modeling of Spatial Vulnerability Applied to a Norflurazon Retrospective Well Study in California, PDF. Troiano, J., J. Marade, and F. Spurlock. J. Environ. Qual. 28:397-403. 1999. Reprinted with the permission of the American Agronomy Society.
The current method used to group and profile sections of land based on soil data is explained in report EH 00-05 and the method to determine depth-to-groundwater estimates for sections of land in report EH 00-02. Both reports are listed below.
- Update of the California Vulnerability Soil Analysis for Movement of Pesticides to Ground Water: October 14, 1999, PDF. Troiano, J., F. Spurlock, and J. Marade. EH 00-05. 2000.
- Procedures for Developing a Depth-to-Ground Water Database, PDF. Spurlock, F. EH 00-02. 2000.
Using this vulnerability analysis, studies have been conducted to identify the pathways of contamination and appropriate management practices.
For content questions or to request documents, contact: