MODELING DOLINE GEOMETRY AND ASSESSING KARST GEOHAZARDS USING GIS

Geographic Information Science (GIS) continues to be an important tool in mapping and spatial analysis of karst terrain. Much of the karst in the United States occurs in the “sunbelt” regions where population growth rates are highest. Doline collapse and flooding sinkholes increasingly pose problems for government, planners, contractors, engineers and geologists. Major obstacles in characterizing karst terrains for spatial analysis include identifying and collecting critical data and categorizing and analyzing large data sets. GIS enables both efficient storage and robust analysis of spatial karst data for multipurpose uses.

The purposes of the GIS-based study are (1) construction of a digital, geo-referenced database for characterizing dolines on the northwestern Highland Rim of Tennessee and its contiguous southwest Pennyroyal Plain of Kentucky; and (2) development of a prediction model to assess the karst hazards of depression flooding and doline collapse. The digital database contains two broad categories of attributes: (1) three-dimensional geometric parameters; and (2) engineering properties of the regolith occupying each bedrock void. Within each category additional sub-attributes were collected, geo-referenced and analyzed. The GIS attributes grouped under the doline geometry category include geo-referenced locations in the state plane coordinate system (SPCS), geologic formation name, depth, perimeter, area, estimated volume, long-axis orientation and length/width ratio of each doline. Variables grouped in the GIS attribute table (dbf) under engineering properties of the regolith include the Unified Soil Classification designation of the subsoil occupying the doline and its Atterberg limits. GIS supports attribute linkages and captures spatial correlations in a way that was previously impossible. In addition, GIS can apply sophisticated methods, such as kriging and co-kriging or sequential Gaussian simulations to develop a systematic sinkhole mapping system capable of predicting potential collapse risks and flooding probabilities.