Numerical Simulation And Uncertainty Analysis Of Jinci Spring Area Karst Groundwater System

Jinci spring,a famous karst spring in Shanxi Province,is one of the important water sources in Taiyuan City.Since the 70's,there has been a great deal of exploitation of karst groundwater in the Jinci spring area,and the exploitation and recharge of karst water resources has been seriously unbalanced,and the groundwater level has continued to decline,the spring was cutoff in April 30,1994 eventually.In this case,to prove the feasibility of the spring water reflow under different recovery schemes,the prediction of Jinci spring reflow time,and the scientificity of the reflow measures are very concerned by the local government and people.In the past,the modflow groundwater flow numerical simulation method is used in the study of Jinci spring reflow,which does not considering the impact of the uncertainties of groundwater model parameters and local turbulence on the simulated prediction results,so there is great uncertainty in the reflow scheme based on these models.The object of this study is Jinci spring karst groundwater system,and analyzes the influence of the uncertainty of the model parameters on the prediction of Jinci spring flow by using GLUE method coupling MODFLOW model.At the same time,the groundwater flow model of Jinci spring area was established by MODFLOW-CFP model,to analyze the influence of local turbulence on the prediction of Jinci spring flow.The main conclusions obtained are as follows:1.In this paper,the MODFLOW model of Jinci spring area is established,and the model can simulate the water level and spring flow of the research area better.Based on this model,the sensitivity analysis of hydrogeological parameters is carried out,and the results show that the flow and water level of Jinci spring are only sensitive to permeability coefficient,compare with water storage coefficient,specific yield.2.Based on the MODFLOW model,the influence of the uncertainty of permeability coefficient on the prediction of Jinci spring flow was analyzed by GLUE method.The results show that the influence of the permeability coefficient uncertainty on the prediction of spring flow become smaller over the time,and under the wet or normal year condition,which the uncertainty of permeability coefficient has little effect on the prediction of spring flow,and the reflow time difference is only two or three years.Under the condition of dry year,the uncertainty of permeability coefficient will lead to twenty or thirty-year reflow time difference.Therefore,if we do not take into account the uncertainty of permeability coefficient,the spring water flow time may be delayed than expected twenty or thirty years,in order to ensure that the implementation of the rehabilitation measures to achieve the expected results,may need to increase the supply and pressure-mining efforts of research areas.3.On the basis of the MODFLOW model,the MODFLOW-CFP model which can reflect the turbulence of groundwater is established,and the sensitivity analysis of CFP parameters of the turbulence cell distribution is carried out.The results show that the turbulent flow often occurs in the regions where with large permeability coefficient,large water level difference between cells,such as the center of the study area and the vent of Jinci spring,and the distribution range vary with the parameters,the larger average pore diameter d,and the higher groundwater temperature T,the wider distribution range of the turbulent cells in the study area.The larger the critical Reynolds number NRe,the smaller the distribution range of turbulence units in the study area.By analyzing the number of turbulent units under each parameter condition,the results show that d and NRe have the same sensitivity to the distribution of turbulence units,which much higher than T.4.In this paper,we select two sets of parameters to analyze the influence of local turbulence on the identification and verification of the model,and the results show that local turbulence affects the water level of the region by influencing the equivalent permeability coefficient,and it will accelerate the water level rise in the area where the water level is rising,and it will accelerate the water level drop in the area where the water level is down As the overall water level of the research area decreased during the identification verification period,as the flow of spring gradually became smaller,the turbulence unit in the vicinity of the spring vent was decreasing,and the equivalent permeability coefficient was relatively larger in the process,and the water level decline rate in the vicinity of the spring vent was slowed down in CFPM2 model.The water level in the vicinity of the spring vent in CFPM2 model is higher than the MODFLOW model at the later stage of the identification verification period,in which the CFPM2(d=10m)model is higher than the CFPM2(d=3m)model.5.The results show that the occurrence of local turbulence will make the spring reflow time advance in all replenishment cases.Under the condition of poor recharge,the reflow time may be much ahead and the flow of spring is larger;When the recharge condition is good,the local turbulent flow will make the forecast spring flow relatively small after a certain period of spring reflow.So,considering the local turbulence influence on springs flow prediction,decreasing reflow engineering efforts to some extent,spring can still occur reflow in the expected time,both to the economic and reflow effect thereby.However,under the condition of good resupply,after a certain period of reflow,it is necessary to strengthen the reflow project to reach the expected flow.This paper provides a new angle for the correct understanding of the karst groundwater fluid state in Jinci spring karst groundwater and the construction of groundwater mode,and provides scientific suggestions for the reflow scheme of Jinci spring and the sustainable exploitation of karst water resources.