Sampling behavioral model parameters for ensemble-based sensitivity analysis using Gaussian process emulation and active subspaces

被引:8
作者
Erdal, Daniel [1 ,2 ]
Xiao, Sinan [3 ]
Nowak, Wolfgang [3 ]
Cirpka, Olaf A. [1 ]
机构
[1] Univ Tubingen, Ctr Appl Geosci, Holderlinstr 12, D-72074 Tubingen, Germany
[2] Tyrens AB, Lilla Badhusgatan 2, S-41121 Gothenburg, Sweden
[3] Univ Stuttgart, Inst Modelling Hydraul & Environm Syst LS3 SimTec, Pfaffenwaldring 5a, D-70569 Stuttgart, Germany
关键词
Global sensitivity analysis; Sampling behavioral models; Gaussian process emulation; Stochastic engine; DIAGNOSE INTEGRATED HYDROLOGY; SUBSURFACE-FLOW; UNCERTAINTY QUANTIFICATION; HYDRAULIC CONDUCTIVITY; DIMENSION REDUCTION; GROUNDWATER; DESIGN; CALIBRATION; SIMULATION; SURROGATES;
D O I
10.1007/s00477-020-01867-0
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
Ensemble-based uncertainty quantification and global sensitivity analysis of environmental models requires generating large ensembles of parameter-sets. This can already be difficult when analyzing moderately complex models based on partial differential equations because many parameter combinations cause an implausible model behavior even though the individual parameters are within plausible ranges. In this work, we apply Gaussian Process Emulators (GPE) as surrogate models in a sampling scheme. In an active-training phase of the surrogate model, we target the behavioral boundary of the parameter space before sampling this behavioral part of the parameter space more evenly by passive sampling. Active learning increases the subsequent sampling efficiency, but its additional costs pay off only for a sufficiently large sample size. We exemplify our idea with a catchment-scale subsurface flow model with uncertain material properties, boundary conditions, and geometric descriptors of the geological structure. We then perform a global-sensitivity analysis of the resulting behavioral dataset using the active-subspace method, which requires approximating the local sensitivities of the target quantity with respect to all parameters at all sampled locations in parameter space. The Gaussian Process Emulator implicitly provides an analytical expression for this gradient, thus improving the accuracy of the active-subspace construction. When applying the GPE-based preselection, 70-90% of the samples were confirmed to be behavioral by running the full model, whereas only 0.5% of the samples were behavioral in standard Monte-Carlo sampling without preselection. The GPE method also provided local sensitivities at minimal additional costs.
引用
收藏
页码:1813 / 1830
页数:18
相关论文
共 67 条
[1]  
[Anonymous], 2008, GLOBAL SENSITIVITY A, DOI DOI 10.1002/9780470725184
[2]  
[Anonymous], 2004, Sensitivity Analysis in Practice: A Guide to Assessing Scientific Models, DOI [10.1111/j.1467-985X.2005.35813.x, DOI 10.1002/0470870958]
[3]  
Aquanty Inc, 2015, HYDR US MAN
[4]   A review of surrogate models and their application to groundwater modeling [J].
Asher, M. J. ;
Croke, B. F. W. ;
Jakeman, A. J. ;
Peeters, L. J. M. .
WATER RESOURCES RESEARCH, 2015, 51 (08) :5957-5973
[5]   Diagnostics for Gaussian Process Emulators [J].
Bastos, Leonardo S. ;
O'Hagan, Anthony .
TECHNOMETRICS, 2009, 51 (04) :425-438
[6]  
Bect J., 2017, STK SMALL MATLAB OCT
[7]   Hierarchical adaptive experimental design for Gaussian process emulators [J].
Busby, Daniel .
RELIABILITY ENGINEERING & SYSTEM SAFETY, 2009, 94 (07) :1183-1193
[8]   An improved adaptive kriging-based importance technique for sampling multiple failure regions of low probability [J].
Cadini, F. ;
Santos, F. ;
Zio, E. .
RELIABILITY ENGINEERING & SYSTEM SAFETY, 2014, 131 :109-117
[9]   Active learning with statistical models [J].
Cohn, DA ;
Ghahramani, Z ;
Jordan, MI .
JOURNAL OF ARTIFICIAL INTELLIGENCE RESEARCH, 1996, 4 :129-145
[10]   Time-dependent global sensitivity analysis with active subspaces for a lithium ion battery model [J].
Constantine, Paul G. ;
Doostan, Alireza .
STATISTICAL ANALYSIS AND DATA MINING, 2017, 10 (05) :243-262