SEMI-AUTOMATED ANALYSIS OF SINKHOLE DISTRIBUTION IN THE LEVELS OF WEST VIRGINIA

Sinkholes, including swallets, serve as major recharge points for water, solutes, and sediments in many karst aquifers. Because they are the most visual karst features, they have been frequently investigated in terms of size, shape, and distribution. The availability of topographic data in digital (GIS) form provides opportunities to analyze large populations of sinkholes, but there are difficulties with automating such analyses. A semi-computerized approach was used to evaluating characteristics of the Big Levels and Little Levels of Southeastern West Virginia, which are located in the Greenbrier River Basin. These areas are limestone uplands that cover portions of three counties within the Appalachian Plateau, and which are home to over 2000 caves, including seven of the longest in North America. Digital line graphs (hypsography) for fifteen, 7.5-minute quadrangles were analyzed for this study using ESRI's ArcGIS.

The creation of the initial sinkhole data set used the hypsography attribute field which designates contour lines as depressions. These depression contours are used to create polygon layers showing total spatial distribution. A numbering system designated the depth of individual sinkholes and showed the distribution of sinkholes of a particular depth. The sinkhole data set was also used to determine relationships between sinkholes and bedrock, hydrology, and elevation. Problems occurring with this method are documented as sources for possible errors in similar studies.

The results of this study showed about 3500 sinkholes in the Levels of West Virginia. An overall sinkhole density of 0.05 square meters of sinkhole per every one square meter of land surface was determined. Analyses of the hydrology of the area showed six percent of the sinkholes occurring within 100 meters of streams as well as 304 ponded sinkholes. Sinkholes in this region are found to occur at a mean elevation of 647 meters and no sinkholes were found below an elevation of 482 meters or above 896 meters. This data was then combined in order to find regions in which sinkholes are most likely to occur, and therefore have an effect on the surrounding environment. The semi-automated analysis of this region provides a useful, wide-scale evaluation regarding the size, shape, and distribution of sinkholes recharging the carbonate aquifer.