On the applicability of surrogate‐based Markov chain Monte Carlo‐Bayesian inversion to the Community Land Model: Case studies at flux tower sites

Huang, Maoyi; Ray, Jaideep; Hou, Zhangshuan; Ren, Huiying; Liu, Ying; Swiler, Laura Painton

doi:10.1002/2015jd024339

Cited by 29 publications

(34 citation statements)

References 72 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this study, we tested emulator calibration with number of parameters that are comparable to previous studies, if not higher (Ray et al, 2015, Huang et al, 2016Gong and Duan, 2017). However, running the emulator can also become infeasible,…”

Section: Approximation Error Vs Clock-timementioning

confidence: 79%

See 1 more Smart Citation

Linking big models to big data: efficient ecosystem model calibration through Bayesian model emulation

Fer

Kelly²,

Andrews

et al. 2018

Preprint

View full text Add to dashboard Cite

Abstract. Data-model integration plays a critical role in assessing and improving our capacity to predict ecosystem dynamics. Similarly, the ability to attach quantitative statements of uncertainty around model forecasts is crucial for model assessment and interpretation and for setting field research priorities. Bayesian methods provide a rigorous data assimilation framework for these applications, especially for problems with multiple data constraints. However, the Markov Chain Monte Carlo (MCMC) techniques underlying most Bayesian calibration can be prohibitive for computationally-demanding models a n d large data sets. We describe an alternative method, Bayesian model emulation of sufficient statistics, that can approximate the full joint posterior density, is more amenable to parallelization, and provides an estimate of parameter sensitivity. Analysis involved informative priors constructed from a meta-analysis of the primary literature, and introduced novel approaches to the specification of both model and data uncertainties, including bias and autocorrelation corrections on multiple data streams. We report the integration of this method within an ecological workflow management software, Predictive Ecosystem Analyzer (PEcAn), and its application and validation with two process-based terrestrial ecosystem models: SIPNET and ED2. In a test against a synthetic dataset, the emulator was able to retrieve the true parameter values. A comparison of the emulator approach to standard "bruteforce" MCMC involving multiple data constraints showed that the emulator method was able to constrain the faster and simpler SIPNET model's parameters with comparable performance to the bruteforce approach, but reduced computation time by more than two orders of magnitude. The emulator was then applied to calibration of the ED2 model, whose complexity precludes standard (bruteforce) Bayesian data assimilation uncertainty around their predictions. Performance metrics showed increased agreement between model predictions and data.Our study furthers efforts toward reducing model uncertainties showing that the emulator method makes it possible to efficiently calibrate complex models. This efficient data assimilation method allows us to conduct more calibration experiments in relatively much shorter times, enabling constraining of numerous models using the expanding amount and types of data.

show abstract

Section: Approximation Error Vs Clock-timementioning

confidence: 79%

“…1) where the model has been run. Previous studies on emulation of biosphere models mostly focused on emulating the model outputs (Kennedy et al, 2008, Ray et al, 2015, Huang et al, 2016. However, comparing model outputs to 'big data' requires emulating a large, nonlinear multivariate output space.…”

Section: Introductionmentioning

confidence: 99%

Linking big models to big data: efficient ecosystem model calibration through Bayesian model emulation

Fer

Kelly²,

Andrews

et al. 2018

Preprint

View full text Add to dashboard Cite

show abstract

“…In this study, we focused on calibrating process-based mechanistic simulators (ecosystem models) using computationally cheaper emulators. Variations in emulator approach are many and can be found in Jandarov et al (2014), Aslanyan et al (2015), Huang et al (2016), Oakley and Youngman (2017), and the references therein. Here we adopted the version that emulates the likelihood surface with a GP, similar to previous studies including applications with a cosmological likelihood function (Aslanyan et al, 2015), a stochastic natural history model (Oakley and Youngman, 2017), the Hartmann function and a hydrologic model (Wang et al, 2014), and two land surface models (Li et al, 2018).…”

Section: Emulator Constructionmentioning

confidence: 99%

“…Emulators are constructed by interpolating a response surface between the knots where the model has been run. Previous studies on emulation of biosphere models mostly focused on emulating the model outputs (Kennedy et al, 2008;Ray et al, 2015;Huang et al, 2016). However, comparing model outputs to "big data" requires emulating a large, nonlinear multivariate output space.…”

mentioning

confidence: 99%

Linking big models to big data: efficient ecosystem model calibration through Bayesian model emulation

Kelly²,

et al. 2018

View full text Add to dashboard Cite

Abstract. Data-model integration plays a critical role in assessing and improving our capacity to predict ecosystem dynamics. Similarly, the ability to attach quantitative statements of uncertainty around model forecasts is crucial for model assessment and interpretation and for setting field research priorities. Bayesian methods provide a rigorous data assimilation framework for these applications, especially for problems with multiple data constraints. However, the Markov chain Monte Carlo (MCMC) techniques underlying most Bayesian calibration can be prohibitive for computationally demanding models and large datasets. We employ an alternative method, Bayesian model emulation of sufficient statistics, that can approximate the full joint posterior density, is more amenable to parallelization, and provides an estimate of parameter sensitivity. Analysis involved informative priors constructed from a meta-analysis of the primary literature and specification of both model and data uncertainties, and it introduced novel approaches to autocorrelation corrections on multiple data streams and emulating the sufficient statistics surface. We report the integration of this method within an ecological workflow management software, Predictive Ecosystem Analyzer (PEcAn), and its application and validation with two process-based terrestrial ecosystem models: SIPNET and ED2. In a test against a synthetic dataset, the emulator was able to retrieve the true parameter values. A comparison of the emulator approach to standard brute-force MCMC involving multiple data constraints showed that the emulator method was able to constrain the faster and simpler SIPNET model's parameters with comparable performance to the brute-force approach but reduced computation time by more than 2 orders of magnitude. The emulator was then applied to calibration of the ED2 model, whose complexity precludes standard (brute-force) Bayesian data assimilation techniques. Both models are constrained after assimilation of the observational data with the emulator method, reducing the uncertainty around their predictions. Performance metrics showed increased agreement between model predictions and data. Our study furthers efforts toward reducing model uncertainties, showing that the emulator method makes it possible to efficiently calibrate complex models.

show abstract

“…Several global optimization methods have been applied in land surface modeling. For example, Sun et al (2013) and Huang et al (2016) used Markov chain Monte Carlo (MCMC) sampling to calibrate hydrological parameters in the Community Land Model (CLM); the MCMC methods have also been applied for CLM-crop model calibration in Zeng et al (2013), and for terrestrial ecosystem model (TEM) optimization in Ricciuto et al (2008Ricciuto et al ( , 2011 and Safta et al (2015). Trudinger et al (2007) and Fox et al (2009) used Kalman filters for TEM parameter estimation.…”

Section: Introductionmentioning

confidence: 99%

Calibration of the E3SM Land Model Using Surrogate‐Based Global Optimization

Ricciuto

Stoyanov

et al. 2018

J Adv Model Earth Syst

View full text Add to dashboard Cite

Calibration of the Energy Exascale Earth System Model (E3SM), land model (ELMv0) is challenging because of its model complexity, strong model nonlinearity, and significant computational requirements. Therefore, only a limited number of simulations can be allowed in any attempt to find a near‐optimal solution within an affordable time. The goal of this study is to calibrate some of the ELMv0 parameters to improve model projection of carbon fluxes. We propose a computationally efficient global optimization procedure using sparse‐grid based surrogates. We first use advanced sparse grid (SG) interpolation to construct a surrogate system of the ELMv0, and then calibrate the surrogate model in the optimization process. As the surrogate model is a polynomial whose evaluation is fast, it can be efficiently evaluated a sufficiently large number of times in the optimization, which facilitates the global search. We calibrate eight parameters against five years of net ecosystem exchange, total leaf area index, and latent heat flux data from the U.S. Missouri Ozark flux tower. The calibrated model is then used for predicting the three variables in the following 4 years. The results indicate that an accurate surrogate model can be created for the ELMv0 with a relatively small number of SG points, i.e., a few ELMv0 simulations that can be fully parallel. And, the application of the optimized parameters leads to a better model performance and a higher predictive capability than the default parameter values in the ELMv0.

show abstract

On the applicability of surrogate‐based Markov chain Monte Carlo‐Bayesian inversion to the Community Land Model: Case studies at flux tower sites

Cited by 29 publications

References 72 publications

Linking big models to big data: efficient ecosystem model calibration through Bayesian model emulation

Linking big models to big data: efficient ecosystem model calibration through Bayesian model emulation

Linking big models to big data: efficient ecosystem model calibration through Bayesian model emulation

Calibration of the E3SM Land Model Using Surrogate‐Based Global Optimization

Contact Info

Product

Resources

About