Multiobjective inverse parameter estimation for modelling vadose zone water movement

Abstract

Inverse modelling techniques for estimating unsaturated soil hydraulic parameters have become increasingly common in the past two decades. In contrast to single-objective parameter estimation which yields a single set of "best fit" parameters, multiobjective parameter estimation results in a number of Pareto optimal solutions which allow the analysis of the trade-off between different, sometimes conflicting, model objectives. In this study, modelling tools for identification of Pareto optimal sets of vadose zone water transport parameters are presented utilizing the numerical water and solute transport model HYDRUS-1D. Root-mean-square error (RMSE) values are calculated to measure the fit of the simulated and observed pressure head data at three different depths at a vadose zone of volcanic origin in New Zealand.