Files
Abstract
This work is an interdisciplinary research involving the fields of statistics and hydrology. In hydrology, rainfall-runoff models give a wide spectrum of output series and calibration procedures require a significant amount of time. Calibration of these hydrologic models especially is a challenging task since the input parameters of these models are often unknown and correspond to physical properties that are difficult to measure. In statistics, model parameter estimation (calibration) problem simplifies to finding an inverse solution of a computer model that generates time series output. In this research, we focus on solving the inverse problem for hydrologic time series and, thus, calibrating the computer model. We propose a modified history matching approach for calibrating rainfall-runoff models efficiently. We present the methodology and illustrate the application of the algorithm using both synthetic and field data (one simulation study and two case studies). We calculate several goodness-of-fit statistics to assess the performance of the modified history matching algorithm. The results demonstrate that the proposed approach improved model performance by 30 % and 11 % in the case studies of compartment model and SWAT model, respectively.