Simulation of regional land subsidence in the southern Yangtze Delta

Abstract

Investigation of the deformation characteristics of individual hydrostratigraphic units is the key to construct a regional land subsidence model. All of 12 hydrostratigraphic units in the study area were discussed throughout. On the basis of the measured data of groundwater level, five kinds of changing patterns of groundwater level were deduced and the relationship between the deformation characteristics of aquifer units and the corresponding changing patterns of groundwater level was discussed. The study area is 1.7 10 4 km 2 , where the geological condition is complex. The changing patterns of groundwater level the hydrostratigraphic units have experienced vary from site to site and from time to time. Consequently, the deformation characteristics of units are sophisticated. An identical hydrostratigraphic unit may exhibit different deformation characteristics, such as elasticity, elasto-plasticity, visco-elasticity, and visco-elasto-plasticity, at different sites or during different periods, not to mention the different units. The existing models are difficult to describe the complex visco-elasto-plastic constitutive law under the condition of land subsidence. So the Merchant's model was modified to depict the visco-elasto-plastic behavior of units. Then a three-dimensional flow model with variable parameters and a vertical one-dimensional subsidence model were constructed and coupled. The coupled model was applied in simulating land subsidence in the southern Yangtze Delta and a satisfactory result was obtained. The simulation results show that the new coupled model can depict the complex geological conditions and describe the developing process of land subsidence very well in the southern Yangtze Delta. The new model can surely be used to predict land subsidence in the future, which is very helpful to taking measurements to control land subsidence. regional land subsidence model, deformation characteristics of soil, creep, couple, numerical simulation Land subsidence, which may result from many factors, is a geological hazard that causes ground elevation to decline slowly. Excessive groundwater withdrawal is a main cause for land subsidence. In China, seventeen provinces (cities) have encountered land subsidence due to groundwater withdrawal. The total subsidence area has reached 90000 km 2 , which has resulted in huge economic loss. In order to control land subsidence effecttively and reduce the damage, and to ensure the safety of high-speed highway and railway, and the sustainable development and reasonable planning, it is currently in urgent need to solve the problem of constructing a regional land subsidence model to predict the developing trend of land subsidence. The cones of depression and the bowls of subsidence have joined together, thus a regional land subsidence often involves an area of 10000 30000 km 2 or more. The previous land subsidence models in China and other countries often covered dozens of or one to two hundred square kilometers and included only one or two hydrostratigraphic units with

Three-dimensional numerical modeling of land subsidence in Shanghai, China

“…The deformation characteristics of the sedimentary layers vary because of differences in lithology, thickness, applied stress, and stress history, which has been accurately described in the literature (Xue et al 2008;Ye et al 2011Ye et al , 2012. A short summary of the most important information is provided here.…”

“…Comprehensive studies on numerical modeling of land subsidence have been conducted by researchers from Nanjing University, China, and the Shanghai Institute of Geological Survey since 2000. A regional land subsidence model with a three-dimensional (3D) groundwater flow module and a 1D subsidence module embodying complex visco-elasto-plastic deformation characteristics with variable parameters, was developed and applied to simulate and predict land subsidence in Shanghai (Ye 2004;Ye et al 2005Ye et al , 2011Xue et al 2008;Shi et al 2008;Wu et al 2010Wu et al , 2102. The 3D groundwater flow model was used to calculate the time-dependent hydraulic heads in the sedimentary sequence, and the 1D non-linear subsidence model considering delayed drainage of pore water was used to compute the time-dependent compaction of the sedimentary layers and the cumulative land subsidence.…”

“…Zero flux on the basement boundary and Dirichlet conditions on the top and lateral boundaries are prescribed in the flow model. The prescribed hydraulic heads on the top and lateral boundaries were obtained from the larger-extent groundwater flow model developed for Shanghai city to simulate the period 1961-2005(Xue et al 2008). A Dirichlet condition with fixed bottom boundary and a traction-free surface are assumed in the geomechanical model.…”

Three dimensional numerical modeling of land subsidence in Shanghai

“…The researchers from the Nanjing University, China, and the Shanghai Institute of Geological Survey have conducted comprehensive studies on numerical modeling of land subsidence since 2000. A regional land subsidence model, composed of a three dimensional (3-D) groundwater flow module and a 1-D subsidence module embodying complex visco-elasto-plastic deformation characteristics with variable parameters, was developed and applied to simulate and predict land subsidence in Shanghai (Ye, 2004;Ye et al, 2005;Xue et al, 2008;Shi et al, 2008;Wu et al, 2010;Ye et al, 2011Ye et al, , 2012. The 3-D groundwater flow model is used to calculate the time-dependent hydraulic heads in the sedimentary sequence, and the 1-D subsidence model is used to compute the time-dependent compaction of the sedimentary layers and the cumulative land subsidence.…”

Large differential land subsidence and earth fissures in Jiangyin, China

“…Theoretically, the long term land subsidence in this context consists of three stages, namely elastic deformation, primary and secondary consolidations. In a microscope, the subsidence may demonstrate a cyclic rebound due to the seasonal fluctuation of groundwater level (Riley 1969;Xue et al 2008;Wang et al 2009d).…”