A Shuffled Complex Evolution Metropolis algorithm for optimization and uncertainty assessment of hydrologic model parameters

Jasper A. Vrugt, Hoshin Vijai Gupta, Willem Bouten, Soroosh Sorooshian

Research output: Contribution to journalArticle

795 Citations (Scopus)

Abstract

Markov Chain Monte Carlo (MCMC) methods have become increasingly popular for estimating the posterior probability distribution of parameters in hydrologic models. However, MCMC methods require the a priori definition of a proposal or sampling distribution, which determines the explorative capabilities and efficiency of the sampler and therefore the statistical properties of the Markov Chain and its rate of convergence. In this paper we present an MCMC sampler entitled the Shuffled Complex Evolution Metropolis algorithm (SCEM-UA), which is well suited to infer the posterior distribution of hydrologic model parameters. The SCEM-UA algorithm is a modified version of the original SCE-UA global optimization algorithm developed by Duan et al. [1992]. The SCEM-UA algorithm operates by merging the strengths of the Metropolis algorithm, controlled random search, competitive evolution, and complex shuffling in order to continuously update the proposal distribution and evolve the sampler to the posterior target distribution. Three case studies demonstrate that the adaptive capability of the SCEM-UA algorithm significantly reduces the number of model simulations needed to infer the posterior distribution of the parameters when compared with the traditional Metropolis-Hastings samplers.

Original languageEnglish (US)
JournalWater Resources Research
Volume39
Issue number8
StatePublished - Aug 2003

Fingerprint

hydrologic models
uncertainty
Markov chain
samplers
Markov processes
sampler
Monte Carlo method
Monte Carlo methods
probability distribution
Global optimization
Merging
Probability distributions
distribution
metropolis
parameter
Uncertainty
simulation models
Sampling
case studies
sampling

Keywords

  • Automatic calibration
  • Hydrologic models
  • Markov Chain Monte Carlo
  • Parameter optimization
  • Proposal distribution
  • Uncertainty assessment

ASJC Scopus subject areas

  • Environmental Science(all)
  • Environmental Chemistry
  • Aquatic Science
  • Water Science and Technology

Cite this

A Shuffled Complex Evolution Metropolis algorithm for optimization and uncertainty assessment of hydrologic model parameters. / Vrugt, Jasper A.; Gupta, Hoshin Vijai; Bouten, Willem; Sorooshian, Soroosh.

In: Water Resources Research, Vol. 39, No. 8, 08.2003.

Research output: Contribution to journalArticle

@article{95e83b4494194415a3e653c00c63dc8f,
title = "A Shuffled Complex Evolution Metropolis algorithm for optimization and uncertainty assessment of hydrologic model parameters",
abstract = "Markov Chain Monte Carlo (MCMC) methods have become increasingly popular for estimating the posterior probability distribution of parameters in hydrologic models. However, MCMC methods require the a priori definition of a proposal or sampling distribution, which determines the explorative capabilities and efficiency of the sampler and therefore the statistical properties of the Markov Chain and its rate of convergence. In this paper we present an MCMC sampler entitled the Shuffled Complex Evolution Metropolis algorithm (SCEM-UA), which is well suited to infer the posterior distribution of hydrologic model parameters. The SCEM-UA algorithm is a modified version of the original SCE-UA global optimization algorithm developed by Duan et al. [1992]. The SCEM-UA algorithm operates by merging the strengths of the Metropolis algorithm, controlled random search, competitive evolution, and complex shuffling in order to continuously update the proposal distribution and evolve the sampler to the posterior target distribution. Three case studies demonstrate that the adaptive capability of the SCEM-UA algorithm significantly reduces the number of model simulations needed to infer the posterior distribution of the parameters when compared with the traditional Metropolis-Hastings samplers.",
keywords = "Automatic calibration, Hydrologic models, Markov Chain Monte Carlo, Parameter optimization, Proposal distribution, Uncertainty assessment",
author = "Vrugt, {Jasper A.} and Gupta, {Hoshin Vijai} and Willem Bouten and Soroosh Sorooshian",
year = "2003",
month = "8",
language = "English (US)",
volume = "39",
journal = "Water Resources Research",
issn = "0043-1397",
publisher = "American Geophysical Union",
number = "8",

}

TY - JOUR

T1 - A Shuffled Complex Evolution Metropolis algorithm for optimization and uncertainty assessment of hydrologic model parameters

AU - Vrugt, Jasper A.

AU - Gupta, Hoshin Vijai

AU - Bouten, Willem

AU - Sorooshian, Soroosh

PY - 2003/8

Y1 - 2003/8

N2 - Markov Chain Monte Carlo (MCMC) methods have become increasingly popular for estimating the posterior probability distribution of parameters in hydrologic models. However, MCMC methods require the a priori definition of a proposal or sampling distribution, which determines the explorative capabilities and efficiency of the sampler and therefore the statistical properties of the Markov Chain and its rate of convergence. In this paper we present an MCMC sampler entitled the Shuffled Complex Evolution Metropolis algorithm (SCEM-UA), which is well suited to infer the posterior distribution of hydrologic model parameters. The SCEM-UA algorithm is a modified version of the original SCE-UA global optimization algorithm developed by Duan et al. [1992]. The SCEM-UA algorithm operates by merging the strengths of the Metropolis algorithm, controlled random search, competitive evolution, and complex shuffling in order to continuously update the proposal distribution and evolve the sampler to the posterior target distribution. Three case studies demonstrate that the adaptive capability of the SCEM-UA algorithm significantly reduces the number of model simulations needed to infer the posterior distribution of the parameters when compared with the traditional Metropolis-Hastings samplers.

AB - Markov Chain Monte Carlo (MCMC) methods have become increasingly popular for estimating the posterior probability distribution of parameters in hydrologic models. However, MCMC methods require the a priori definition of a proposal or sampling distribution, which determines the explorative capabilities and efficiency of the sampler and therefore the statistical properties of the Markov Chain and its rate of convergence. In this paper we present an MCMC sampler entitled the Shuffled Complex Evolution Metropolis algorithm (SCEM-UA), which is well suited to infer the posterior distribution of hydrologic model parameters. The SCEM-UA algorithm is a modified version of the original SCE-UA global optimization algorithm developed by Duan et al. [1992]. The SCEM-UA algorithm operates by merging the strengths of the Metropolis algorithm, controlled random search, competitive evolution, and complex shuffling in order to continuously update the proposal distribution and evolve the sampler to the posterior target distribution. Three case studies demonstrate that the adaptive capability of the SCEM-UA algorithm significantly reduces the number of model simulations needed to infer the posterior distribution of the parameters when compared with the traditional Metropolis-Hastings samplers.

KW - Automatic calibration

KW - Hydrologic models

KW - Markov Chain Monte Carlo

KW - Parameter optimization

KW - Proposal distribution

KW - Uncertainty assessment

UR - http://www.scopus.com/inward/record.url?scp=1542757127&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=1542757127&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:1542757127

VL - 39

JO - Water Resources Research

JF - Water Resources Research

SN - 0043-1397

IS - 8

ER -