GIS Assessment Methods

The source water assessment areas determine the boundary for assessment calculations for each public water intake. As previously discussed, the delineation of the assessment areas varied depending upon the type of water system being evaluated. For instance, Arkansas water systems using wells as a water source were referred to as groundwater systems and given an assessment radius that varied from 240 ft to 14,000 ft depending upon recommendations by the ADH water system specialists.

In order to complete a source water assessment, a number of environmental characteristics are typically measured, computed, or obtained for each of the assessment areas. This information was more easily calculated by harnessing the capabilities of a Geographic Information System (GIS). A detailed description of the actual data summary methodology that was used within ArcView 3.2 is contained in Appendix-B of this document.

The "GeoAccounting" tools used for the Assessment Calculations

After the source water assessment areas were delineated, a variety of geospatial analysis operations were performed; occasionally, in batch operations to calculate summary statistics for the assessment area(s). These "geostatistics" would later be incorporated into the Department of Health's assessment algorithms for computation of each public water system's susceptibility to contamination.

Geocoding the Potential Sources of Contamination

The process of Geocoding the addresses of local businesses that contain, use or store potentially harmful chemicals and materials was very beneficial to the collection of the "non-regulated" PSOC locations (ie. dry cleaners, hardware stores, golf courses, etc).

For Arkansas' source water assessment, a business database was obtained from Acxiom Corporation InfoBase Products (www.acxiom.com). Each InfoBase record contained a business name, address and business type classification (Standard Industrial Codes, SIC). By using geocoding software such as MapMarker (MapInfo Corp.) to tag each business address with an X, Y coordinate, we were able to map those businesses classifications that pose a potential threat to a water source. The MapMarker software was selected because we could easily control the spatial accuracy of the geocoded PSOCs. Therefore, any address not geocoded to an actual street-segment within the TIGER/1998 street data layer (approximately +/- 167-ft) could be removed or identified as a "questionable location" for the source water assessment. The source water team determined that any address not geocoded to the correct street segment would be removed from the potential contamination (PSOC) list.

Proximity-to-Intake Determination

The distance of each PSOC to the Source Water intake is also of great concern to the Arkansas Department of Health and CAST calculated these euclidean distances for every PSOC within each assessment area.

Area calculations

The area and percent coverage for various landscape characteristics, such as Land Use, Slope, Geology, and General Soil characteristics were also calculated for each assessment area. These geospatial operations were performed much like "cookie-cutting". By using the assessment areas as the "cutter" and any other map layer as the "dough", the geostatistics could be summarized for each public water intake. The area and percent coverage calculations of each map layer were effective for quantifying the surface conditions found within each of the source water assessment areas.

Buffer analysis

Buffering operations are considered basic spatial analysis, but are of utmost importance to source water assessments. The Arkansas Department of Health's (non-graphical) database contained unique identification codes for each PWS intake, as well as, a radius distance (in feet) for the groundwater assessment areas. Although these radii varied from 240 to over 14,000 feet, the ArcView3.2 buffering wizard enabled the selection of the "radius column" and therefore, the assessment areas for 1300 groundwater systems were calculated in a batch operation.

Point-in-Polygon analysis

Whether extracting the elevation of each intake from a Digital Elevation Model (DEM) or adding additional attribution to the database of source water intakes (eg. county names), we were able to take advantage of "spatial correlation" within the GIS. We also updated existing database attributes by overlaying one theme, such as county population, annual precipitation, or elevation with another map theme (such as intake locations). This geospatial tool was very useful for value-adding existing source water data layers.

Reclassification of Attributes

The conversion of the attributes within various GIS layers to another was very beneficial to the source water assessment project, for instance, when the soil attributes were converted to a "weighting factor" for the susceptibility models developed by the Arkansas Department of Health. This basic operation was used on several occasions to make the attributes of the Arkansas geospatial data more user friendly to the project team.

Map and Report Generation

Example Map and Report (Click to view an example of a map)

CAST created an Avenue script in ArcView3.2 for map and report production, which enabled the mass production of over 1400 maps that were mailed out during the spring of 1999 for data validation by the public water system managers. By automating this process, the GIS Analyst were able to easily produce the (8.5"x11") maps and individual PSOC reports for each Public Water System intake. During a three-month period, approximately 70% of the PSOC locations were evaluated for errors and corrected, before the final summary reports and statistics were calculated. This entire process of map and report production, mail-out, field-verification, and return of the paper maps to the ADH, took slightly over 4 months. More importantly, the process assured the source water project team that 70% of their assessment data had been "verified" by those individuals responsible for the integrity of their water system. All of the spatial and/or attribute errors that were identified during the field verification process were corrected at the Center for Advanced Spatial Technologies. Additional attribution was added to each PSOC record, so that the Arkansas Department of Health officials could identify those individual data points that were updated during the spring of 1999.