Analysis and mapping of mining subsidence and underground voids collapse risk using GIS and the weights-of-evidence model for the abandoned Ichmoul mine, Algeria

Abstract

The present study concerns the assessment of mining subsidence and old underground works collapse hazard of the abandoned Ichmoul mine (Batna, Algeria) using the geographic information system (GIS) and the weight of evidence method (WofE). After the identification of collapse cases on the site and developing a spatial database for old underground works, topography and geology, the weight of evidence model was applied to calculate the weight of each relevant factor. Five main factors controlling or related to subsidence and underground voids collapse were determined from the probability analysis of existing cases; the underground voids depth, the distance between the underground voids, the landforms slope, the lithology and the distance between the fractures. Conditional independence tests were performed for the choice of factors. For the analysis of the subsidence and underground void collapse spatial hazard mapping, the contrast values, W+ and W- of the each factor evaluation were analyzed. The analysis results were validated using the curve ROC (AUC) with a past collapse case. For all the factors used, the area under the (ROC) curve showed 92 % accuracy. The results obtained can be used for the prevention of mining subsidence and underground voids collapse risks and the mine site rehabilitation.