Effective and efficient algorithm for multiobjective optimization of hydrologic models

By Jasper A. Vrugt, Hoshin V. Gupta, Luis A. Bastidas, Willem Bouten, Soroosh Sorooshian. 2003

677 citations recorded by the source corpus.

Practical experience with the calibration of hydrologic models suggests that any single‐objective function, no matter how carefully chosen, is often inadequate to properly measure all of the characteristics of the observed data deemed to be important. One strategy to circumvent this problem is to define several optimization criteria (objective functions) that measure different (complementary) aspects of the system behavior and to use multicriteria optimization to identify the set of nondominated, efficient, or Pareto optimal solutions. In this paper, we present an efficient and effective Markov Chain Monte Carlo sampler, entitled the Multiobjective Shuffled Complex Evolution Metropolis (MOSCEM) algorithm, which is capable of solving the multiobjective optimization problem for hydrologic models. MOSCEM is an improvement over the Shuffled Complex Evolution Metropolis (SCEM‐UA) global optimization algorithm, using the concept of Pareto dominance (rather than direct single‐objective function evaluation) to evolve the initial population of points toward a set of solutions stemming from a stable distribution (Pareto set). The efficacy of the MOSCEM‐UA algorithm is compared with the original MOCOM‐UA algorithm for three hydrologic modeling case studies of increasing complexity.

Subfields: Global and Planetary Change, Ocean Engineering, Water Science and Technology

View Paper: https://www.semanticscholar.org/paper/9c9fedad94cc06f68baa05d986f34eb4785e7020

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Global and Planetary Change

In Ocean Engineering

In Water Science and Technology

Citations in the corpus

Model identification for hydrological forecasting under uncertainty (2005, methodology): “Gupta et al. (1998) proposed the use of a multi-objective (MO) framework as a way towards addressing this problem (for MO approaches see also for example Yang and Haan 1991; Vrugt et al. 2003a).”
Reconciling theory with observations: elements of a diagnostic approach to model evaluation (2008, background): “ideas have been further codified in the argument that all model identification problems are inherently multicriteria in nature (Gupta et al., 1998; Boyle et al., 2001; Wagener et al., 2001, 2003a, b; Vrugt et al., 2003), and that the challenge lies in (a) proper selection of measures of model perfor”
On the use of spatial regularization strategies to improve calibration of distributed watershed models (2010, methodology): “The method has been widely applied in calibration of hydrological models [e.g., Vrugt et al., 2003; Leplastrier et al., 2002; Xia et al., 2002; Demarty et al., 2005].”
Generalized likelihood uncertainty estimation (GLUE) using adaptive Markov Chain Monte Carlo sampling (2008, methodology): “This correlation plot is consistent with previous results presented in Vrugt et al. (2003b).”
Model Parameter Estimation Experiment (MOPEX): An overview of science strategy and major results from the second and third workshops (2006, methodology): “Examples include the Generalized Likelihood Uncertainty Estimation (GLUE) by Beven and Binley (1992), the Markov Chain Monte Carlo (MCMC) Metropolis scheme by Kuczera and Parent (1998) and the Shuffled Complex Evolution Method Metropolis (SCEM) scheme by Vrugt et al. (2003) .”
Analysing trade-offs in resource and labour allocation by smallholder farmers using inverse modelling techniques: A case-study from Kakamega district, western Kenya (2007, methodology): “Experiments conducted using standard synthetic multiobjective test problems have shown that the final population provides a fairly uniform approximation to the Pareto solution space (Vrugt et al., 2003).”
An Intercomparison of Sampling Methods for Uncertainty Quantification of Environmental Dynamic Models (2015, background): “Furthermore, multi-physics dynamic models also demand multiobjective approach in model calibration (Vrugt et al., 2003; Liu et al., 2005).”
Dynamically dimensioned search algorithm for computationally efficient watershed model calibration (2007, methodology): “Further evidence of the stature of the SCE algorithm is the incorporation of SCE into advanced algorithms for uncertainty analysis [ Vrugt et al. , 2003a, 2003b] and multiobjective optimization [ Vrugt et al. , 2003a; Yapo et al. , 1998].”
Sensitivity analysis and parameter estimation for distributed hydrological modeling: potential of variational methods (2009, background): “These techniques enable the global optimization of single or multiple objectives and very often characterise the uncertainty affecting model parameters (Beven and Bin-ley, 1992; Duan et al., 1992; Kuczera and Parent, 1998; Yapo et al., 1997; Vrugt et al., 2003a,b).”
Model Parameter Estimation Experiment (MOPEX): An overview of science strategy and major results from the second and third workshops (2006, background): “The question of how to estimate model parameters has been receiving increasing attention from the hydrology and land surface modeling community (Franks and Beven, 1997; Bastidas et al., 1999; Gupta et al., 1999; Duan et al., 2001; Duan et al., 2003; Jackson et al., 2003; Wagener et al., 2003).”
Ensemble‐based assimilation of discharge into rainfall‐runoff models: A comparison of approaches to mapping observational information to state space (2009, methodology): “A large number of methods have been developed for this purpose, for example the Parameter Estimation (PEST) method [Doherty, 2001], Bayesian recursive parameter estimation [Thiemann et al., 2001], the Shuffled Complex Evolution (SCE-UA) algorithm [Duan et al., 1994; Yapo et al., 1998; Vrugt et al.,”