Abstract: The correlation-based localized ensemble data assimilation method can be used to estimate model parameters when there is no physical distance between observation and parameter. The estimation accuracy of model parameters is impacted by the way to obtain the tapering factor,which is termed tapering effect,jointly determined by tapering function and correlation coefficient. To investigate how tapering effect impacts the estimate accuracy of hydraulic conductivity,we estimate the spatially random conductivity field of a two-dimensional porous-confined aquifer by considering a variety of combinations of ensemble size,types of correlation coefficient(including Kendall,Spearman,and Pearson),and tapering function. The results show that(i) the Pearson-based localization approaches perform best among the three types of correlation coefficients considered,followed by the Spearman-based ones; and (ii) the tapering effect,which is characterized as the composition of elliptic equation and Gaspari-Cohn function(hyperbolic tangent function or exponential function),leads to overall better performance than the others. The results and the framework of the correlation-based localized iterative ensemble smoother proposed in this study can provide important references for estimating hydrogeology parameters.