Comparison of the Generalized Likelihood Uncertainty Estimation and Markov Chain Monte Carlo Methods for Uncertainty Analysis of the ORYZA_V3 Model

Tan, Junwei; Cao, Jiyue; Cui, Yuanlai; Duan, Qingyun; Gong, Wei

doi:10.2134/agronj2018.05.0336

Cited by 17 publications

(17 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Tan et al [ 84 ] contrasted results of the GLUE and MCMC approaches, assessing uncertainties of nine parameters of a crop model. The authors did not address, however, an explicit specification of parameter interactions according to the posterior joint distribution.…”

Section: Discussionmentioning

confidence: 99%

A Bayesian Approach to the Estimation of Parameters and Their Interdependencies in Environmental Modeling

Albert

Callies

Toussaint

2022

Entropy

View full text Add to dashboard Cite

We present a case study for Bayesian analysis and proper representation of distributions and dependence among parameters when calibrating process-oriented environmental models. A simple water quality model for the Elbe River (Germany) is referred to as an example, but the approach is applicable to a wide range of environmental models with time-series output. Model parameters are estimated by Bayesian inference via Markov Chain Monte Carlo (MCMC) sampling. While the best-fit solution matches usual least-squares model calibration (with a penalty term for excessive parameter values), the Bayesian approach has the advantage of yielding a joint probability distribution for parameters. This posterior distribution encompasses all possible parameter combinations that produce a simulation output that fits observed data within measurement and modeling uncertainty. Bayesian inference further permits the introduction of prior knowledge, e.g., positivity of certain parameters. The estimated distribution shows to which extent model parameters are controlled by observations through the process of inference, highlighting issues that cannot be settled unless more information becomes available. An interactive interface enables tracking for how ranges of parameter values that are consistent with observations change during the process of a step-by-step assignment of fixed parameter values. Based on an initial analysis of the posterior via an undirected Gaussian graphical model, a directed Bayesian network (BN) is constructed. The BN transparently conveys information on the interdependence of parameters after calibration. Finally, a strategy to reduce the number of expensive model runs in MCMC sampling for the presented purpose is introduced based on a newly developed variant of delayed acceptance sampling with a Gaussian process surrogate and linear dimensionality reduction to support function-valued outputs.

show abstract

Section: Discussionmentioning

confidence: 99%

A Bayesian Approach to the Estimation of Parameters and Their Interdependencies in Environmental Modeling

Albert

Callies

Toussaint

2022

Entropy

View full text Add to dashboard Cite

show abstract

“…That seemed not to be the case in our application. Tan et al (2019) contrasted results of the GLUE and MCMC approaches, assessing uncertainties of nine parameters of a crop model. The authors did not address, however, an explicit specification of parameter interactions according to the posterior joint distribution.…”

Section: Discussionmentioning

confidence: 99%

Parameter Dependences Arising from Calibration of a Riverine Diatom Model - Representation in Terms of Posterior Conditional Distributions

Callies

2021

Preprint

View full text Add to dashboard Cite

We address the analysis and proper representation of posterior dependence among parameters obtained from model calibration. A simple water quality model for the Elbe River (Germany) is referred to as an example. The joint posterior distribution of six model parameters is estimated by Markov Chain Monte Carlo sampling based on a quadratic likelihood function. The estimated distribution shows to which extent model parameters are controlled by observations, highlighting issues that cannot be settled unless more information becomes available. In our example, some vagueness occurs due to problems in distinguishing between the effects of either growth limitation by lack of silica or a temperature dependent algal loss rate. Knowing such indefiniteness of the model structure is crucial when the model is to be used in support of management options. Bayesian network technology can be employed to convey this information in a transparent way.

show abstract

“…Generally, independent and uniform priors were used in the previous studies, and the model errors between simulations and observations were assumed to be normally distributed, which has been proved to be very effective in calculating the posterior parameter distributions for a crop model (Tan et al., 2019). Thus, the likelihood function in this study can be written as:

L (boldY |normalθ) = \prod_{i = 1}^{n} \frac{1}{\sqrt{2 {π σ}_{i}^{2}}} \exp ((), - \frac{{((), Y_{i} - O_{i})}^{2}}{2 σ_{i}^{2}})

where Y i and O i are the i th simulation and observation; n is the number of observations; σ i is the variance of model error, and a constant value is assumed here.…”

Section: Methodsmentioning

confidence: 99%

“…Thus, the values of these water stress parameters in the model may be changed for current rice cultivars growing in different environments. However, in the previous studies, only crop parameters in the model have been systematically studied for sensitivity analysis and uncertainty analysis (Tan et al, 2016(Tan et al, , 2019, whereas no similar studies have been found for the drought stress parameters. Therefore, the objectives of this paper are (a) to explore the sensitivities of drought stress parameters in the ORYZA (v3) model on yield under different soil conditions and fertilizer applications, (b) to quantify the ranges and distributions of drought stress parameters, and (c) to evaluate the model performance of yield simulations under drought stress conditions.…”

Section: Core Ideasmentioning

confidence: 99%

“…The MCMC method has been widely used in a range of hydrological models and crop models for uncertainty analysis (Tan et al, 2019;Wang et al, 2017) and parameter estimation (Lv et al, 2016). This method can calculate posterior distributions for model parameters by taking into account prior knowledge of parameter values based on the Bayes' theorem:…”

Section: Uncertainty Analysis For the Drought Stress Parametersmentioning

confidence: 99%

See 1 more Smart Citation

Global sensitivity analysis and uncertainty analysis for drought stress parameters in the ORYZA (v3) model

et al. 2021

Self Cite

View full text Add to dashboard Cite

Drought stress parameters in crop models have received little attention in the literature. In this study, a global sensitivity analysis on yield simulation was conducted for the drought stress parameters in the ORYZA (v3) model, and sensitive parameters were identified for double-season rice (Oryza sativa L.) (early rice and late rice) by using the Extended FAST method in different scenarios involving two N applications, two soil types, and four depths of plough soil layer. An adaptive metropolis algorithm (MCMC-AM) was used to quantify the uncertainties of these parameters with yield observations under various drought stress conditions. Based on the estimated parameter values, model performance of yield simulation was evaluated for drought stress conditions. Results showed that in a light drought stress situation, the upper limit factor for transpiration decline (FSWTD) was extremely sensitive at low N application, whereas the upper threshold for leaf expansion became more sensitive at high N application. In a moderate drought stress situation, the upper threshold for leaf rolling had high sensitivity under all conditions except for the late rice at low N application. In a heavy drought stress situation, FSWTD always had strong influence on yield, whereas the impacts of other parameters varied under different scenarios. Soil types and plough soil depth had considerable effects on the parameter sensitivities depending on fertilizer application and drought stress level. The empirical posterior distributions of these drought stress parameters for late rice showed very different shapes and variation ranges from that for early rice, and large uncertainties were presented. The yield simulations with a finely calibrated model were overestimated under heavy drought stress, especially for the late rice, which was grown in the dry season. The results highlighted that more attention should be paid for the drought stress parameters in crop models.

show abstract

Comparison of the Generalized Likelihood Uncertainty Estimation and Markov Chain Monte Carlo Methods for Uncertainty Analysis of the ORYZA_V3 Model

Cited by 17 publications

References 42 publications

A Bayesian Approach to the Estimation of Parameters and Their Interdependencies in Environmental Modeling

A Bayesian Approach to the Estimation of Parameters and Their Interdependencies in Environmental Modeling

Parameter Dependences Arising from Calibration of a Riverine Diatom Model - Representation in Terms of Posterior Conditional Distributions

Global sensitivity analysis and uncertainty analysis for drought stress parameters in the ORYZA (v3) model

Contact Info

Product

Resources

About