Files
Abstract
Failing septic systems contribute to compromised surface and ground water quality. Counties and regulatory agencies seeking to develop strategies to mitigate this potential source of pollution require, but typically lack, a digitized septic system database which can be logistically expensive to develop. To support that effort, a toolkit to locate and date septic systems was developed and tested in Jackson County, Georgia. Multi-temporal, high-resolution aerial imagery was classified using a supervised support vector machine algorithm to identify buildings. The building classification resulted in a 79% accuracy, and the results were geoprocessed with necessary vector layers in a GIS framework to develop Septic System Automated Location Tool (SSALT). The validation of SSALT with geocoded septic permits for Jackson county showed that 83% parcels were correctly identified to be on septic systems, 7% were incorrectly assigned on septic systems while 10% of parcels identified on septic systems did not contain a building.