Multiobjective calibration with Pareto preference ordering: An application to rainfall‐runoff model calibration

Abstract

[1] Automatic calibration routines for hydrologic models with multiple objective capabilities are becoming increasingly popular due to advances in computational power, population-based optimization techniques, and the recognition that a single performance measure such as the root-mean-square error is no longer sufficient to characterize the complex behavior of the catchment. However, as more objective functions are included in the calibration, the number of Pareto-optimal solutions as well as the number of ''near'' Pareto-optimal parameter sets increases. The calibration problem quickly becomes a decision-making problem. In the practical sense, users of automatic calibration routines have to face the task of selecting a set of suitable model parameters from the numerous Pareto-optimal sets. A new method of automatic calibration is proposed, which combines an effective optimization routine, based on multiobjective genetic algorithms, and Pareto preference ordering. In this case, Pareto-optimal points that are also Pareto-optimal in different subspace combinations of the objective functions space are preferred. The proposed method is used to calibrate the MIKE11/NAM rainfall-runoff model for a Danish catchment. The results indicated that the method is able to sieve through the numerous Pareto-optimal solutions and select a small number of preferred solutions. This is extremely useful to modelers who typically are required to provide the best estimated parameter sets with good overall model performance.

Trade-offs and responsiveness of the single-layer urban canopy parametrization in WRF: An offline evaluation using the MOSCEM optimization algorithm and field observations

“…Designed for the automatic calibration of hydrological models, the Multiobjective Shuffled Complex Evolution Metropolis (MOSCEM) algorithm of Vrugt et al (2003) provides a systematic and unbiased tool for such a purpose. Of particular interest is its ability to: (1) objectively sample the entire parameter space rather than a discrete predefined set of values, and (2) optimize with respect to several criteria (multi-objective optimization), thereby providing some insight into the trade-offs which control the modelled SEB (multi-objective optimization identifies situations where improving one criteria is possible only at the expense of another: Gupta et al, 1998;Khu and Madsen, 2005;Confesor and Whittaker, 2007;Shafii and De Smedt, 2009). The tradeoff surfaces, composed of all parameter values leading to an optimum compromise in the performance of the modelled fluxes, are referred to as Pareto fronts to highlight the nonuniqueness of the solution .…”

Multiobjective Automatic Parameter Calibration of a Hydrological Model

“…The two objectives were simultaneously optimized; they were not aggregated in this study. In the field of hydrological model calibration, the NSGA-II, Multiobjective Complex Evolution (MOCOM) algorithm and the Multiobjective Shuffled Complex Evolution Metropolis (MOSCEM) algorithm are widely used [8,[46][47][48][49][50][51][52][53][54][55][56][57][58][59]. A set of Pareto optimal solutions (non-dominated parameter sets) was obtained from the model.…”

A constraint-based search algorithm for parameter identification of environmental models

“…The multi-objective approach seeks to identify parameter sets that simultaneously provide optimal performance for different aspects of system response (Gupta et al, 1998;Boyle et al, 2000Boyle et al, , 2001. This can include constraining the model to reproduce multiple system fluxes and state variables such as runoff, evaporation, groundwater levels or tracer concentrations (e.g., Gupta et al, 1999;Bastidas et al, 1999;Freer et al, 2002;McDonnell, 2002, 2013;Khu and Madsen, 2005;Fenicia et al, 2008;Winsemius et al, 2008;Birkel et al, 2011;Hrachowitz et al, 2013).…”