Abstract: The risk factors causing karst collapse in Baiyun district are recognized. The hierarchical structure model for evaluating karst collapse risk is formulated. Weight values of basic risk factors resulting in karst collapse regarded as a risk event is calculated by analytical hierarchy process. Karst collapse gradation division is defined as the most likely subsidence area Ⅰ, the more likely subsidence area Ⅱ, the likely subsidence area Ⅲ, the lower likely subsidence area Ⅳ and the stable area Ⅴ. 5 grades of each element whose influential degree on karst collapse is considered, are defined in accordance with karst collapse gradation division. To quantify risk evaluation of karst collapse, the single factor collapse probability pi is introduced and the probability calculation model of karst collapse is established on the basis of Bernoulli trials. Eight basic risk factors including overburden thickness, karst development degree, groundwater depth, bedrock lithology, overburden structure and bottom plate soil properties, groundwater exploitation intensity, underground activity intensity, ground surface construction load and vibration are chosen as the parameters involving in calculation of subsidence probability. Then a narrowly-defined grade criterion of karst collapse is determined, providing a reference for subsidence gradation division. It is indicated that Jiangcun, Gulougang, Quanxi, Xihu and Shegang are attributed to the most likely subsidence area, the more likely and likely subsidence areas distribute mostly in Jianggao town and Yuanshijing town. The risk gradation division results are adaptable to regional subsidence history statistical data and hazard degree assessment of subsidence, so as to prove the effectiveness of the comprehensive evaluation method in this context.

Key words: Baiyun district, karst collapse, risk evaluation, analytic hierarchy process, probabilistic calculation, risk gradation division