An Intercomparison of Sampling Methods for Uncertainty Quantification of Environmental Dynamic Models

Gong, Wei; Duan, Qingyun; Li, J. D.; Wang, C.; Di, Zhenhua; Ye, A. Z.; Miao, Chiyuan; Dai, Yongjiu

doi:10.3808/jei.201500310

Cited by 24 publications

(27 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Given that MC-based sampling strategies can be computationally expensive and sometimes unaffordable for computationally demanding models, other sampling strategies have been developed and improved over the last several decades. Of these, Latin hypercube sampling (LHS;McKay et al, 1979) has been most commonly used for uncertainty and sensitivity analysis in the field of water and environmental modelling (Hossain et al, 2006;Gong et al, 2015;Higdon et al, 2013;Sheikholeslami and Razavi, 2017). The LHS approach offers a sampling strategy that can significantly reduce the sample size without compromising the accuracy of uncertainty estimation compared to the MC sampling approach (Iman and Conover, 1980;Iman and Helton, 1988;McKay et al, 1979).…”

Section: Introductionmentioning

confidence: 99%

Characterizing uncertainty in the hydraulic parameters of oil sands mine reclamation covers and its influence on water balance predictions

Alam

Barbour

Huang

2020

Hydrol. Earth Syst. Sci.

View full text Add to dashboard Cite

One technique to evaluate the performance of oil sands reclamation covers is through the simulation of long-term water balance using calibrated soil-vegetationatmosphere transfer models. Conventional practice has been to derive a single set of optimized hydraulic parameters through inverse modelling (IM) based on short-term ( < 5-10 years) monitoring datasets. This approach is unable to characterize the impact of variability in the cover properties. This study utilizes IM to optimize the hydraulic properties for 12 soil cover designs, replicated in triplicate, at Syncrude's Aurora North mine site. The hydraulic parameters for three soil types (peat cover soil, coarse-textured subsoil, and lean oil sand substrate) were optimized at each monitoring site from 2013 to 2016. The resulting 155 optimized parameter values were used to define distributions for each parameter/soil type, while the progressive Latin hypercube sampling (PLHS) method was used to sample parameter values randomly from the optimized parameter distributions. Water balance models with the sampled parameter sets were used to evaluate variations in the maximum sustainable leaf area index (LAI) for five illustrative covers and quantify uncertainty associated with long-term water balance components and LAI values. Overall, the PLHS method was able to better capture broader variability in the water balance components than a discrete interval sampling method. The results also highlight that climate variability dominates the simulated variability in actual evapotranspiration and that climate and parameter uncertainty have a similar influence on the variability in net percolation.

show abstract

Section: Introductionmentioning

confidence: 99%

Characterizing uncertainty in the hydraulic parameters of oil sands mine reclamation covers and its influence on water balance predictions

Alam

Barbour

Huang

2020

Hydrol. Earth Syst. Sci.

View full text Add to dashboard Cite

show abstract

“…In our previous studies Gan et al 2014;Gong and Duan 2017;Gong et al 2015;Gong et al 2016a;Gong et al 2016b), an uncertainty quantification (UQ) framework was developed to quantify the uncertainty of large, complex dynamic system models, such as land-surface, weather, and climate models, which include many physical processes and cost a substantial amount of computational resources to run. The uncertainty quantification framework includes the following steps: (1) using parameter screening to reduce the number of adjustable parameters, (2) building a surrogate model to emulate the response surfaces of the original model to the variation in the adjustable parameters, and (3) running the surrogate model with the sensitive parameters to optimize the original model.…”

Section: Introductionmentioning

confidence: 99%

How parameter specification of an Earth system model of intermediate complexity influences its climate simulations

Shi

Gong

Duan

et al. 2019

Prog Earth Planet Sci

View full text Add to dashboard Cite

Earth system models (ESMs) consist of parameterization schemes based on one's perception of how the Earth system functions. A typical ESM contains a large number of parameters (i.e., the constants and exponents in the parameterization schemes) whose specification can have a significant impact on an ESM's simulation capabilities. Sensitivity analyses (SA) is an important tool for assessing how parameter specification influences model simulations. In this study, we used an Earth system model of intermediate complexity (EMIC)-LOVECLIM as an example to illustrate how SA methods can be used to identify the most sensitive parameters that control the simulations of several key global water and energy cycle variables, including global annual mean absolute surface air temperature (T G ), precipitation and evaporation over the land and over the oceans (P L , P O , E L , E O ), and land runoff (R L ). We also demonstrate how judiciously specifying model parameters can improve the simulations of those variables. Three SA methods MARS, RF, and sparse PCE-based Sobol' method were used to evaluate a pool of 25 adjustable parameters chosen from land, atmosphere, and ocean components of LOVECLIM and their results were intercompared to ensure robustness of the results. It is found that with different parameter specification, T G can vary from 10 to 20°C, and the values of P L , P O , E L , and E O can change by more than 100%. An interesting observation is that the value of R L vary from 13,000 to 35,000 km 3 , far below the observed climatological value of 40,000 km 3 , indicating a model structural deficiency in representing land runoff by LOVECLIM which must be corrected to obtain more reasonable global water budgets. We also note that parameter sensitivities are significantly different at different latitudes. Finally, we showed that global water and energy cycle simulations can be significantly improved by even a crude automatic parameter tuning, indicating that parameter optimization can be a viable way to improve ESM climate simulations. The results from this study should help us to understand the parameter uncertainty of a full-scale ESM.

show abstract

“…In addition to oil spill simulation, optimization is also desired to provide decision support under changing environmental conditions (Huang et al, 1996;Huang and Cao, 2011;Gong et al, 2016). Zhong and You (2011) developed a multi objecttive linear model for operational cleanup schedules and coastal protection plans during an oil spill event.…”

Section: Introductionmentioning

confidence: 99%

ASOC: A Novel Agent-Based Simulation-Optimization Coupling Approach-Algorithm and Application in Offshore Oil Spill Responses

Li¹,

Chen²,

Li³

et al. 2016

J ENVIRON INFORM

View full text Add to dashboard Cite

ABSTRACT. The efficiency of offshore oil spill response not only relies on an efficaciously global decision/planning in devices combination and allocation, but also depends on the timely control for response devices (e.g., skimmers and booms). However, few studies have reported on such decision framework with a timely integration of global planning and operation control to support offshore oil spill response. This study developed an agent-based simulation-optimization approach to provide sound decisions for device combination and allocation during offshore oil spill recovery in a fast, dynamic and cost-efficient manner under uncertain conditions. Meanwhile, the proposed approach aimed at providing operation control schemes for different devices, reflecting the site conditions, and correspondingly adjusting the global planning in a real-time manner. Such functions would be extremely helpful in the harsh environments prevailing in offshore Newfoundland. In the case study, the developed approach was applied to determine the allocation of 3 responding vessels in collecting spilled oil at 7 locations. The routes of the responding vessels for response operation were optimized and reflected by the principle agent-based programming. Furthermore, several oil weathering processes (e.g., evaporation and dispersion) were also taken into account in the optimization. The modeling results indicated that a minimal timeframe of 21 hours was needed for vessel allocation and recovery operation, leading to an oil recovery rate of 90%. By taking evaporation and dispersion into account, the optimal time window was 18 hours, leading to an oil recovery rate of 75%, an evaporation rate of 12%, and a dispersion rate of 3%. The proposed approach can timely and effectively support the optimal allocation of devices, the control of operation, and the real-time adjustment of global decision making for oil recovery under dynamic conditions.

show abstract

An Intercomparison of Sampling Methods for Uncertainty Quantification of Environmental Dynamic Models

Cited by 24 publications

References 50 publications

Characterizing uncertainty in the hydraulic parameters of oil sands mine reclamation covers and its influence on water balance predictions

Characterizing uncertainty in the hydraulic parameters of oil sands mine reclamation covers and its influence on water balance predictions

How parameter specification of an Earth system model of intermediate complexity influences its climate simulations

ASOC: A Novel Agent-Based Simulation-Optimization Coupling Approach-Algorithm and Application in Offshore Oil Spill Responses

Contact Info

Product

Resources

About