Fully Coupled Simulation And Uncertainty Analysis Of Surface Water And Groundwater On Quality In Hunhe River Basin

In the hydrologic cycle of watershed scale,surface water and groundwater play an important role,and they transform each other and have close hydraulic relation.It is difficult to describe the movement and distribution of surface water and groundwater in the actual system by analyzing surface water or groundwater alone.In order to realize the overall function of surface water and groundwater system,solving the problems of rational allocation of water resources and comprehensive treatment of water pollution,it is an inevitable trend to study the coupling simulation of surface water and groundwater on watershed scale.In the current research on coupling simulation of surface water and groundwater,which most of the researches involved at home and abroad adopt the method of iterative coupling to connect surface water and groundwater system through exchange flux.In addition,most of the researches only conduct coupling simulation study on the regional scale of surface water and groundwater flow,and do not involve the coupling simulation of surface water and groundwater on quality of the whole basin.Taking the whole basin as the operating platform,the water quality coupling simulation model is further established on the basis of the coupling simulation of groundwater and surface water,which has very important theoretical and practical significance to the comprehensive treatment of water pollution in the whole Hunhe River basin.Fully coupled simulation of surface water and groundwater on quality has great uncertainty because of the random variation of model parameters.It is difficult to evaluate the reliability of simulation results without consideration.Therefore,it is necessary to analyze the uncertainty of coupled simulation of surface water and groundwater on quality.In this process,in order to reduce the input dimension of the alternative model,using local sensitivity analysis method to screen out the model parameters with high sensitivity.Thus,it is easier and more efficient to establish the alternative model on the premise of ensuring the accuracy of the alternative model.This paper takes the whole Hunhe River basin as the research area,and the spatial and temporal distribution characteristics of surface water and groundwater quality in Hunhe River basin are simulated and predicted,which through the integrated application of the coupled simulation model of flow and quality of surface water and groundwater,local sensitivity analysis,Latin hypercube sampling,alternative model and monte-carlo method.Then uncertainty analysis of water quality coupling simulation was carried out,and analyzing the influence of the output results on random changes of model parameters.Firstly,on the basis of field investigation,dynamic monitoring and qualitative analysis,understanding the topography,meteorology,hydrology,hydrogeological conditions and land use in the study area,and determining the migration and distribution characteristics of surface water and groundwater.Based on the conceptual model of flow and water quality of?groundwater and surface water,the coupling mathematical simulation model of flow and water quality of?groundwater and surface water is established.In addition,HygroGeoSphere technology system is used to solve the simulation model by parallel synchronous solution,which the mathematical model of surface water,the mathematical model of groundwater and the coupling model are solved at the same time.Based on the synchronous dynamic monitoring data of surface water and groundwater,coupled simulation model of surface water and groundwater on flow and water quality are calibrated and checked.Then,the local sensitivity analysis method is used to calculate the output results of the coupled model,and the parameters with high sensitivity are selected as sensitive parameters.Taking sensitive parameters and fertilizer application amount as random variables and other parameters as deterministic variables,and the Latin hypercubic sampling method was used to sample the random variables and obtain the input sample data set.By running the coupling simulation model of groundwater quality of surface water and using the input-output data set as training samples,an alternative model based on radial basis function was established,and the accuracy of the alternative model was calculated and analyzed.Finally,under the premise of considering the random change of model parameters,the Latin hypercube sampling method is adopted to extract input samples according to the set scenario.By running the alternative model,the response changes of solute mass in groundwater and surface water were predicted.Then according to the monte-carlo simulation method,the uncertainty analysis of the forecast results is carried out.Combined with the ecological environmental requirements and the results of uncertainty analysis in hunhe river basin,the risk assessment was carried out on the condition that the solute quality of surface water and groundwater exceeded the ecological threshold.After the above research,the following conclusions are drawn:(1)Based on the data of synchronous dynamic monitoring of surface water and groundwater,fully coupled simulation of surface water and groundwater flow and quality in Hunhe River basin was calibrated and verified,and the measured and calculated values of surface water and groundwater fit well with each other,so it shows that the coupled simulation model of surface water and groundwater on quality in Hunhe River basin accords with the actual situation of the study area.(2)The results of local sensitivity analysis show that evaporation coefficient,permeability coefficient and dispersion are the most sensitive and take them as sensitive parameters for the next calculation.(3)Using the alternative model of the Kernel Extreme learning machines which based on radial basis function can reduce a lot of calculation load on the premise of ensuring the accuracy of calculation.(4)On the premise of considering the random changes of model parameters,the uncertainty analysis of the prediction results of different input schemes is carried out,and the results show that:the random change of model parameters has an impact on the uncertainty of the prediction results of the coupled simulation model,while the random change of fertilizer application rate has an impact on the solute quality in groundwater and surface water,which does not have an impact on the uncertainty of the output results of the model.