Model emulation and moment-independent sensitivity analysis: An application to environmental modelling

Borgonovo, Emanuele; Castaings, William; Tarantola, Stefano

doi:10.1016/j.envsoft.2011.06.006

Cited by 140 publications

(85 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…A question that meta-model users need to address in any meta-modeling practice is whether an exact fit to the set of design sites or an approximate fit, possibly with smoothing capabilities, is required. Therefore, it is essential to assess the accuracy of a meta-model for prediction before it can be used for SA studies (Stephens et al, 2011;Borgonovo et al, 2012). Despite advances in meta-modeling based SA in many fields, the uncertainty assessment of meta-modeling based SA should be further explored in the future.…”

Section: Meta-modeling Methodsmentioning

confidence: 99%

“…Song et al (2012aSong et al ( ,b, 2013 combined the Sobol' method and response surface model (RSM) approach (RSMSobol'; e.g., the SVM, multivariate adaptive regression splines (MARS)) to estimate parameter sensitivity for hydrological models, involving the Xinanjiang and distributed time-variant gain models (DTVGM). Borgonovo et al (2012) pointed out that the meta-model allows an accurate estimation of density-based sensitivity measures when the main structural features of the original model are captured.…”

Section: Meta-modeling Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Global sensitivity analysis in hydrological modeling: Review of concepts, methods, theoretical framework, and applications

Song

Zhang

Zhan

et al. 2015

Journal of Hydrology

474

273

View full text Add to dashboard Cite

s u m m a r y Sensitivity analysis (SA) aims to identify the key parameters that affect model performance and it plays important roles in model parameterization, calibration, optimization, and uncertainty quantification. However, the increasing complexity of hydrological models means that a large number of parameters need to be estimated. To better understand how these complex models work, efficient SA methods should be applied before the application of hydrological modeling. This study provides a comprehensive review of global SA methods in the field of hydrological modeling. The common definitions of SA and the typical categories of SA methods are described. A wide variety of global SA methods have been introduced to provide a more efficient evaluation framework for hydrological modeling. We review, analyze, and categorize research into global SA methods and their applications, with an emphasis on the research accomplished in the hydrological modeling field. The advantages and disadvantages are also discussed and summarized. An application framework and the typical practical steps involved in SA for hydrological modeling are outlined. Further discussions cover several important and often overlooked topics, including the relationship between parameter identification, uncertainty analysis, and optimization in hydrological modeling, how to deal with correlated parameters, and time-varying SA. Finally, some conclusions and guidance recommendations on SA in hydrological modeling are provided, as well as a list of important future research directions that may facilitate more robust analyses when assessing hydrological modeling performance.

show abstract

Section: Meta-modeling Methodsmentioning

confidence: 99%

Section: Meta-modeling Methodsmentioning

confidence: 99%

Global sensitivity analysis in hydrological modeling: Review of concepts, methods, theoretical framework, and applications

Song

Zhang

Zhan

et al. 2015

Journal of Hydrology

474

273

View full text Add to dashboard Cite

show abstract

“…Thus, SA can be defined as an application of qualitative or quantitative tools to study how the variation in the output of a model can be attributed to the uncertainty in the input information provided [24]. As a result, it provides a valuable method to determine which input parameters are important (factor prioritisation) and which factors are non-influential (factor fixing) and rank them in order of importance [25,26]. The latter can be significant since low-impact modelled outputs may be converted to fixed values or dropped to simplify the model, reducing the required computing power.…”

Section: Introductionmentioning

confidence: 99%

Extending the Global Sensitivity Analysis of the SimSphere model in the Context of its Future Exploitation by the Scientific Community

Petropoulos

Ireland

Griffiths

et al. 2015

Water

View full text Add to dashboard Cite

In today's changing climate, the development of robust, accurate and globally applicable models is imperative for a wider understanding of Earth's terrestrial biosphere. Moreover, an understanding of the representation, sensitivity and coherence of such models are vital for the operationalisation of any physically based model. A Global Sensitivity Analysis (GSA) was conducted on the SimSphere land biosphere model in which a meta-modelling method adopting Bayesian theory was implemented. Initially, effects of assuming uniform probability distribution functions (PDFs) for the model inputs, when examining sensitivity of key quantities simulated by SimSphere at different output times, were examined. The development of topographic model input parameters (e.g., slope, aspect, and elevation) were derived within a Geographic Information System (GIS) before implementation within the model. The effect of time of the simulation on the sensitivity of previously examined outputs was also analysed. Results showed that simulated outputs were significantly influenced by changes in topographic input parameters, fractional vegetation cover, vegetation height and surface moisture availability in agreement with previous studies. Time of model output simulation had a significant influence on the absolute values OPEN ACCESSWater 2015, 7 2102 of the output variance decomposition, but it did not seem to change the relative importance of each input parameter. Sensitivity Analysis (SA) results of the newly modelled outputs allowed identification of the most responsive model inputs and interactions. Our study presents an important step forward in SimSphere verification given the increasing interest in its use both as an independent modelling and educational tool. Furthermore, this study is very timely given on-going efforts towards the development of operational products based on the synergy of SimSphere with Earth Observation (EO) data. In this context, results also provide additional support for the potential applicability of the assimilation of spatial analysis data derived from GIS and EO data into an accurate modelling framework.

show abstract

“…Global methods are often based on the analysis of the output variance and are known as ANOVA (ANalysis Of VAriance) techniques [9,10,7]. More recently, sensitivity moment-independent methods have been used, where emulation model is used to compute densitybased sensitivity measure [11]. The emulator is the one of [12].…”

Section: Introductionmentioning

confidence: 99%

Sensitivity study of dynamic systems using polynomial chaos

Sandoval

Anstett-Collin

Basset

2012

Reliability Engineering & System Safety

View full text Add to dashboard Cite

Global sensitivity has mainly been analyzed in static models, though most physical systems can be described by differential equations. Very few approaches have been proposed for the sensitivity of dynamic models and the only ones are local. Nevertheless, it would be of great interest to consider the entire uncertainty range of parameters since they can vary within large intervals depending on their meaning. Other advantage of global analysis is that the sensitivity indices of a given parameter are evaluated while all the other parameters can be varied. In this way, the relative variability of each parameter is taken into account, revealing any possible interactions. This paper presents the global sensitivity analysis for dynamic models with an original approach based on the polynomial chaos (PC) expansion of the output. The evaluation of the PC expansion of the output is less expensive compared to direct simulations. Moreover, at each time instant, the coefficients of the PC decomposition convey the parameter sensitivity and then a sensitivity function can be obtained. The PC coefficients are determined using non-intrusive methods. The proposed approach is illustrated with some well-known dynamic systems.

show abstract

Model emulation and moment-independent sensitivity analysis: An application to environmental modelling

Cited by 140 publications

References 53 publications

Global sensitivity analysis in hydrological modeling: Review of concepts, methods, theoretical framework, and applications

Global sensitivity analysis in hydrological modeling: Review of concepts, methods, theoretical framework, and applications

Extending the Global Sensitivity Analysis of the SimSphere model in the Context of its Future Exploitation by the Scientific Community

Sensitivity study of dynamic systems using polynomial chaos

Contact Info

Product

Resources

About