Study on the statistical prediction model of karst groundwater table in the Jiaozuo coalmine

The Jiaozuo coalmine, 1 300 km2 in area, situates in the southeast edge of the Taihangshan uplift in China. The Ordovician limestone aquifer and the Carboniferous limestone aquifer are separated into the upper and the lower aquifer in the mining area by bauxitic mudstone. Because of the impact of the late period tectonic movement, the bauxitic mudstone lost the waterproof function locally, leading the upper and the lower limestone aquifers get some hydraulic connections, which not only aggravate the disaster accidents, but also increase certain difficulty for the restoration of mine water damage. In view of being no complete groundwater system structure data in the study area, the statistics prediction model and delay model of groundwater table is established on the basis of principles of mass conservation and calculus to discuss the response characteristics of karst groundwater to climate and analyze the main influence factors, and then provide theoretical support for mine water disaster prevention and control. The calculation results show that, when the time delay is 10 months, the correlation coefficient of karst groundwater table and precipitation is up to maximum 0.83, and when the time delay is 2 months or 5 months, the correlation coefficient is only 0.71 and 0.78. The predicting results of groundwater table by prediction model show that the observation hole of the correlation coefficient over 0.90 is up to 60 %, and the observation hole of the correlation coefficient over 0.80 reach 95 %. The model level influence factors analysis shows that the regime of karst groundwater table is most sensitive to rainfall in the coalmine area among the three factors — the rainfall, the mine drainage, and the groundwater evaporation. It is showed that the rainfall recharge response, especially the long-term stability recharge is the main factor influencing the groundwater table. The models, simple and convenient, have better applicability and application prospect.