MAPPING KARST FEATURES FROM LIDAR-DERIVED ELEVATION MODELS IN MONROE COUNTY, WEST VIRGINIA

Monroe County in southeastern West Virginia hosts world-class karst within carbonate units of Mississippian and Ordovician age. The U.S. Geological Survey is mapping the geology of the county and making an inventory of karst features as part of an ongoing water resource investigation. Lidar-derived elevation data acquired in 2016 were used to create a 3m resolution working digital elevation model (DEM), from which surface depressions were identified using a semi-automated workflow in ArcGIS^®. Polygons of closed depressions in carbonate units with areas greater than 100 m² and a minimum depth of 0.2 m were created using a contour-tree identification approach, such that individual depressions were defined along the highest localized closed contour interval of 0.5 m that did not enclose any other separate depressions.

Each depression in the automated inventory was visually checked by a geologist using aerial imagery, lidar-derived imagery, and 3D viewing of the lidar imagery within Google Earth^®. Depressions were systematically checked within a grid of 1.5 km² tiles, and distinguishing features such as modification by human activities or hydrological significance (stream sink, ephemerally ponded, etc.) were noted. Relative confidence in depression identification was determined by whether the depression was visible in the lidar imagery, aerial imagery, or both. Over 2500 total depressions were identified; of these, roughly half (48%) of all depressions were noted with high confidence as visible in both aerial and lidar imagery.

Statistics on elevation, depth, and geometric morphometry of each depression were calculated including perimeter, area, volume, length of major and minor elliptical axes, and azimuth of the major axis. Azimuths of the major elliptical axes were compared against mapped topographic lineaments using rose diagrams. The resulting patterns were similar; the mean azimuth of the major axes of the depressions was 045, and that of the lineaments was 041. The average maximum depth of all depressions was 3.0 m; the average area was 3002 m³. The volume of the largest depression in the study area was 2,971,532 m³ with an area of 282,834 m² and a maximum depth of 25 m. 173 depressions were noted as probable stream sinks (7% of the total depressions) and targeted for field verification.