Multi-objective global optimization for hydrological models

• Published in: Journal of hydrology (Amsterdam) 1998-01, Vol.204 (1/4), p.83-97
• Main Authors: Yapo, PO, Gupta, H V, Sorooshian, S
• Format: Article
• Language: English
• ISSN: 0022-1694

The Multi-Objective Complex Evolution (MOCOM-UA) algorithm, an effective and efficient methodology for solving the multiple-objective global optimization problem of hydrological models, is presented. The formulation and Pareto optimality of the multi-objective calibration problem are discussed then the MOCOM-UA algorithm is explained and illustrated in flow charts. The calibration of the model is described and the importance of population size emphasized. The method was an extension of the Shuffled Complex Evolution global optimization algorithm. It was an advance on the selection of a single objective measure of the distance between the model-simulated output and the data and the selection of a computer-based optimization algorithm to search for the parameter values which minimized that distance. Practical experience suggested a single-objective function was adequate to measure the ways in which the model failed to match the important characteristics of the observed data. Given that some of the latest hydrological models simulated several of the catchment output fluxes, there was a need for effective and efficient multi-objective calibration procedures capable of exploiting all of the useful information about the physical system contained in the measurement data time series. The features and capabilities of the algorithm were illustrated by means of a simple hydrological model calibration study. There are 61 references.