GIS-BASED SPATIAL ANALYSIS FOR THE PREDICTION OF FLOODED AREA IN AN UNDERGROUND LIMESTONE MINE

This study is to predict the possible flooding area in an underground limestone mine using the spatial analysis based in geologic information system (GIS). The dissolved limestone rock mass by the infiltration of ground water or precipitation into bed rock can trigger the safety problem in underground mine.

The study area is Sambo limestone mine in Danyang, South Korea. Field surveys were conducted to obtain information on groundwater flow into the mine drifts, rock mass rating (RMR), Q-system and geological structure. In addition, topographic map and mine drift map of that underground mine were gathered. From literature review and field surveys data, five key factor that may influence a flooding event were selected for the spatial analysis; RMR, Q-system, drift depth, distance from a fault and the surface runoff accumulation. They were analyzed with some spatial analysis such as ordinary kriging, Euclidean distance method and hydrologic analysis and then, the GIS thematic maps were established for a frequency ratio (FR) model.

The FR model was implemented to categorize the statistical classes for each factor. The weight values to construct the flooding occurrence map were determined by the R2 coefficients. The AUC (Area Under Curve) technique revealed that the flooding occurrence map matched well with the previously recorded flooding areas.