Study On Improvement Of Hydraulic Conductivity Inversion Precision Under The Influence Of Correlation Scale Error

The spatial variability of hydraulic conductivity is an important factor affecting the uncertainty of seepage field,so it is necessary to accurately grasp the spatial distribution of hydraulic conductivity of aquifer and effectively describe solute transport in aquifer with scarce borehole data,so as to reduce the uncertainty of description of the nature of aquifer.In order to study the influence of correlation scale error on the inversion precision of the hydraulic conductivity of the aquifer and how to improve the inversion precision effectively.In this paper,the successive linear estimator(SLE)is used to inverse the hydraulic conductivity field of the heterogeneous aquifer based on the scenario simulation.The head observations obtained by scenario simulation were taken as the head observations in the inversion of SLE parameters,and the correlation scale values were changed based on the given correlation scale calculated by scenario simulation to continue the SLE parameters inversion.The spatial distribution of the hydraulic conductivity estimated value is obtained by the SLE parameter inversion,and compared the error between the hydraulic conductivity estimated value and the real value.By increasing the number of observed wells,K field measured values and pumping tests,the relationship between the number of observed wells number,K field measured values volume,pumping tests number and the inversion results error is obtained under the different correlation scale errors conditions.By analyzing and studying the above examples,the following conclusions can be drawn:(1)When the observed head data volume is less,on the basis of the given correlation scale in scenario simulation,reduced the inversion correlation scale value,correlation scale error will affect the inversion result.On the basis of the given correlation scale in scenario simulation,increased the inversion correlation scale value,correlation scale error will not affect the inversion result.(2)The inversion correlation scale value is changed based on the scenario simulation correlation scale value.By increasing the number of observed wells,the volume of K field measured values and the number of pumping tests to reduce the inversion error.It is found that the improvement degree of inversion precision slows down with the increase of data volume.Produce this phenomenon is mainly due to the inversion result error is caused by the correlation scale error and insufficient observed head data volume and numerical calculation error.With the increase of data volume,the errors of inversion results caused by correlation scale error and insufficient observed head data volume are gradually be offset,but the error of numerical computation is more and more apparent.(3)On the basis of the scenario simulation correlation scale value,to reduce the inversion correlation scale.When the correlation scale error is too large,the inversion precision cannot be significantly improved by increasing the observed wells,the K field measured values and the pumping tests.However,on the basis of the scenario simulation correlation scale value,the inversion correlation scale is increased.As the correlation scale error increases,the inversion precision will still be significantly improved by increasing the observed wells,the K field measured values and the pumping tests.Therefore,in practical engineering,no matter which method is adopted,the inversion result will be better if the correlation scale value take larger.(4)On the basis of the scenario simulation correlation scale,when increased the inverse correlation scale,as the correlation scale error increases,the error that can be reduced by increasing the observed wells number,the K field measured values volume and the pumping tests number gradually decreases.(5)The results show that in the aspect of improve the inversion precision,when the same correlation scale error,reduce the inverse correlation scale,the increasing observed wells method is better than the increasing K field measured values method better than the increasing pumping tests method.Increase the inverse correlation scale,the increasing pumping tests method is better than the increasing observed wells method better than the increasing K field measured values method.(6)According to the numerical experiment results,it is suggested to adopt add the pumping tests method when the inversion result precision is high.When the requirement for inversion precision is not particularly high,the method of increasing K field measured values method should be considered.If the inverse correlation scale value reduced too much on the basis of the scenario simulation correlation scale,no matter which method,it is difficult to get the ideal inversion result.