Simple approach to emulating complex computer models for global sensitivity analysis

Stanfill, Bryan; Mielenz, Henrike; Clifford, David; Thorburn, Peter J.

doi:10.1016/j.envsoft.2015.09.011

Cited by 25 publications

(18 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We could fit the full model and then compute this marginal distribution following Stanfill et al (2015). However, an easier and quicker way is to fit a GAM to the (X i , Y j ) "data" where X i and Y j are defined above.…”

Section: The Generalised Additive Modelling Methodsmentioning

confidence: 99%

“…Instead, it is possible to obtain accurate estimates of the first-order SIs by using a GAM to estimate the components of Eq. (1) directly (Stanfill et al, 2015;Strong et al, 2014Strong et al, , 2015b. To compute the ith first-order SI (1 ≤ i ≤ p), we first recognise that taking the expectation of Eq.…”

Section: The Generalised Additive Modelling Methodsmentioning

confidence: 99%

“…Stanfill et al (2015) found that the GAM method was accurate at estimating SIs based on a simple model (three to four parameters) as well as a more complex one (10 parameters). However, if more models of varying complexity and type (e.g.…”

Section: Comparison Of Sensitivity Indicesmentioning

confidence: 99%

“…Given recent interest in applying GAMs to perform GSA (Strong et al, 2015a(Strong et al, , b, 2014, only Stanfill et al (2015) have compared how they perform against other variance-based approaches. The authors found that first-order SIs estimated from the original FAST method were very close to the true values using 600 executions of the model, whereas the GAM approach only required 90-150 model runs.…”

Section: Comparison Of Sensitivity Indicesmentioning

confidence: 99%

See 3 more Smart Citations

Fast sensitivity analysis methods for computationally expensive models with multi-dimensional output

et al. 2018

View full text Add to dashboard Cite

Abstract. Global sensitivity analysis (GSA) is a powerful approach in identifying which inputs or parameters most affect a model's output. This determines which inputs to include when performing model calibration or uncertainty analysis. GSA allows quantification of the sensitivity index (SI) of a particular input -the percentage of the total variability in the output attributed to the changes in that input -by averaging over the other inputs rather than fixing them at specific values. Traditional methods of computing the SIs using the Sobol and extended Fourier Amplitude Sensitivity Test (eFAST) methods involve running a model thousands of times, but this may not be feasible for computationally expensive Earth system models. GSA methods that use a statistical emulator in place of the expensive model are popular, as they require far fewer model runs. We performed an eight-input GSA, using the Sobol and eFAST methods, on two computationally expensive atmospheric chemical transport models using emulators that were trained with 80 runs of the models. We considered two methods to further reduce the computational cost of GSA: (1) a dimension reduction approach and (2) an emulator-free approach. When the output of a model is multi-dimensional, it is common practice to build a separate emulator for each dimension of the output space. Here, we used principal component analysis (PCA) to reduce the output dimension, built an emulator for each of the transformed outputs, and then computed SIs of the reconstructed output using the Sobol method. We considered the global distribution of the annual column mean lifetime of atmospheric methane, which requires ∼ 2000 emulators without PCA but only 5-40 emulators with PCA. We also applied an emulator-free method using a generalised additive model (GAM) to estimate the SIs using only the training runs. Compared to the emulator-only methods, the emulator-PCA and GAM methods accurately estimated the SIs of the ∼ 2000 methane lifetime outputs but were on average 24 and 37 times faster, respectively.

show abstract

Section: The Generalised Additive Modelling Methodsmentioning

confidence: 99%

Section: The Generalised Additive Modelling Methodsmentioning

confidence: 99%

Section: Comparison Of Sensitivity Indicesmentioning

confidence: 99%

Section: Comparison Of Sensitivity Indicesmentioning

confidence: 99%

See 2 more Smart Citations

Fast sensitivity analysis methods for computationally expensive models with multi-dimensional output

et al. 2018

View full text Add to dashboard Cite

show abstract

“…Stanfill et al (2015) found that the GAM the behaviour of the model; although we have found it is a good GSA method for our models…”

mentioning

confidence: 99%

Fast sensitivity analysis methods for computationally expensive models with multidimensional output

Ryan¹,

Wild²,

O’Connor³

et al. 2017

Preprint

View full text Add to dashboard Cite

show abstract

Simulating spatial variability of groundwater table in England and Wales

Rahman

Pianosi

Woods

2023

Hydrological Processes

View full text Add to dashboard Cite

Groundwater is a vital natural resource that affects human life and the environment.Groundwater supplies drinking water, supports irrigation for food production, and helps to sustain flow of water in streams and wetlands during dry periods. Therefore, it is important to assess the variability of groundwater resources and consider it in freshwater management. Numerical modelling is widely used for understanding the variability of groundwater table. However, there is still a lack of large-scale groundwater flow models. For instance, in the United Kingdom, groundwater modelling efforts so far have focused on aquifer to regional scales. A national scale model would aid the quantification of the available groundwater storage and the assessment of the impacts of climate change and anthropogenic activities on this crucial resource. In this study, we evaluate a numerical groundwater flow modelling approach for entire England and Wales. This model operates at 1 km spatial and

show abstract

Simple approach to emulating complex computer models for global sensitivity analysis

Cited by 25 publications

References 28 publications

Fast sensitivity analysis methods for computationally expensive models with multi-dimensional output

Fast sensitivity analysis methods for computationally expensive models with multi-dimensional output

Fast sensitivity analysis methods for computationally expensive models with multidimensional output

Simulating spatial variability of groundwater table in England and Wales

Contact Info

Product

Resources

About