Based on observed data from Tanggu District in Tianjin, a back-propagation neural network (BPNN) model was introduced to predict possible land subsidence due to exploitation of groundwater. According to model estimati...Based on observed data from Tanggu District in Tianjin, a back-propagation neural network (BPNN) model was introduced to predict possible land subsidence due to exploitation of groundwater. According to model estimation under various hypothetical extraction scenarios, patterns of land subsidence at Tanggu District were studied and discussed.The predicted average background land subsidence rate of Tanggu is 9.47 mm/a.The significance of contribution of aquifers to land subsidence descends in order of units Ⅳ,Ⅲ,Ⅴ,Ⅱ.Land subsidence tends to deteriorate with the increase in total extraction rate.展开更多
基金Supported by Tianjin Land Subsidence Controlling Office(No.kJ/095).
文摘Based on observed data from Tanggu District in Tianjin, a back-propagation neural network (BPNN) model was introduced to predict possible land subsidence due to exploitation of groundwater. According to model estimation under various hypothetical extraction scenarios, patterns of land subsidence at Tanggu District were studied and discussed.The predicted average background land subsidence rate of Tanggu is 9.47 mm/a.The significance of contribution of aquifers to land subsidence descends in order of units Ⅳ,Ⅲ,Ⅴ,Ⅱ.Land subsidence tends to deteriorate with the increase in total extraction rate.