An overview of uncertainty quantification techniques with application to oceanic and oil-spill simulations

Mohamed Iskandarani, Shitao Wang, Ashwanth Srinivasan, W. Carlisle Thacker, Justin Winokur, Omar M. Knio

Research output: Contribution to journalArticle

13 Citations (Scopus)

Abstract

We give an overview of four different ensemble-based techniques for uncertainty quantification and illustrate their application in the context of oil plume simulations. These techniques share the common paradigm of constructing a model proxy that efficiently captures the functional dependence of the model output on uncertain model inputs. This proxy is then used to explore the space of uncertain inputs using a large number of samples, so that reliable estimates of the model's output statistics can be calculated. Three of these techniques use polynomial chaos (PC) expansions to construct the model proxy, but they differ in their approach to determining the expansions' coefficients; the fourth technique uses Gaussian Process Regression (GPR). An integral plume model for simulating the Deepwater Horizon oil-gas blowout provides examples for illustrating the different techniques. A Monte Carlo ensemble of 50,000 model simulations is used for gauging the performance of the different proxies. The examples illustrate how regression-based techniques can outperform projection-based techniques when the model output is noisy. They also demonstrate that robust uncertainty analysis can be performed at a fraction of the cost of the Monte Carlo calculation.

Original languageEnglish (US)
Pages (from-to)2789-2808
Number of pages20
JournalJournal of Geophysical Research C: Oceans
Volume121
Issue number4
DOIs
StatePublished - Apr 1 2016

Fingerprint

Oil spills
oil spill
oils
simulation
plumes
regression analysis
output
plume
blowout
expansion
Gaging
Uncertainty analysis
Uncertainty
oil
uncertainty analysis
chaotic dynamics
Gas oils
Chaos theory
horizon
chaos

Keywords

  • Gaussian processes
  • integral plume model
  • polynomial chaos
  • uncertainty quantification

ASJC Scopus subject areas

  • Geochemistry and Petrology
  • Geophysics
  • Earth and Planetary Sciences (miscellaneous)
  • Space and Planetary Science
  • Oceanography

Cite this

An overview of uncertainty quantification techniques with application to oceanic and oil-spill simulations. / Iskandarani, Mohamed; Wang, Shitao; Srinivasan, Ashwanth; Carlisle Thacker, W.; Winokur, Justin; Knio, Omar M.

In: Journal of Geophysical Research C: Oceans, Vol. 121, No. 4, 01.04.2016, p. 2789-2808.

Research output: Contribution to journalArticle

Iskandarani, Mohamed ; Wang, Shitao ; Srinivasan, Ashwanth ; Carlisle Thacker, W. ; Winokur, Justin ; Knio, Omar M. / An overview of uncertainty quantification techniques with application to oceanic and oil-spill simulations. In: Journal of Geophysical Research C: Oceans. 2016 ; Vol. 121, No. 4. pp. 2789-2808.
@article{b4e5fc6779a1487ab6c405fe56285f60,
title = "An overview of uncertainty quantification techniques with application to oceanic and oil-spill simulations",
abstract = "We give an overview of four different ensemble-based techniques for uncertainty quantification and illustrate their application in the context of oil plume simulations. These techniques share the common paradigm of constructing a model proxy that efficiently captures the functional dependence of the model output on uncertain model inputs. This proxy is then used to explore the space of uncertain inputs using a large number of samples, so that reliable estimates of the model's output statistics can be calculated. Three of these techniques use polynomial chaos (PC) expansions to construct the model proxy, but they differ in their approach to determining the expansions' coefficients; the fourth technique uses Gaussian Process Regression (GPR). An integral plume model for simulating the Deepwater Horizon oil-gas blowout provides examples for illustrating the different techniques. A Monte Carlo ensemble of 50,000 model simulations is used for gauging the performance of the different proxies. The examples illustrate how regression-based techniques can outperform projection-based techniques when the model output is noisy. They also demonstrate that robust uncertainty analysis can be performed at a fraction of the cost of the Monte Carlo calculation.",
keywords = "Gaussian processes, integral plume model, polynomial chaos, uncertainty quantification",
author = "Mohamed Iskandarani and Shitao Wang and Ashwanth Srinivasan and {Carlisle Thacker}, W. and Justin Winokur and Knio, {Omar M.}",
year = "2016",
month = "4",
day = "1",
doi = "10.1002/2015JC011366",
language = "English (US)",
volume = "121",
pages = "2789--2808",
journal = "Journal of Geophysical Research: Oceans",
issn = "2169-9275",
publisher = "Wiley-Blackwell",
number = "4",

}

TY - JOUR

T1 - An overview of uncertainty quantification techniques with application to oceanic and oil-spill simulations

AU - Iskandarani, Mohamed

AU - Wang, Shitao

AU - Srinivasan, Ashwanth

AU - Carlisle Thacker, W.

AU - Winokur, Justin

AU - Knio, Omar M.

PY - 2016/4/1

Y1 - 2016/4/1

N2 - We give an overview of four different ensemble-based techniques for uncertainty quantification and illustrate their application in the context of oil plume simulations. These techniques share the common paradigm of constructing a model proxy that efficiently captures the functional dependence of the model output on uncertain model inputs. This proxy is then used to explore the space of uncertain inputs using a large number of samples, so that reliable estimates of the model's output statistics can be calculated. Three of these techniques use polynomial chaos (PC) expansions to construct the model proxy, but they differ in their approach to determining the expansions' coefficients; the fourth technique uses Gaussian Process Regression (GPR). An integral plume model for simulating the Deepwater Horizon oil-gas blowout provides examples for illustrating the different techniques. A Monte Carlo ensemble of 50,000 model simulations is used for gauging the performance of the different proxies. The examples illustrate how regression-based techniques can outperform projection-based techniques when the model output is noisy. They also demonstrate that robust uncertainty analysis can be performed at a fraction of the cost of the Monte Carlo calculation.

AB - We give an overview of four different ensemble-based techniques for uncertainty quantification and illustrate their application in the context of oil plume simulations. These techniques share the common paradigm of constructing a model proxy that efficiently captures the functional dependence of the model output on uncertain model inputs. This proxy is then used to explore the space of uncertain inputs using a large number of samples, so that reliable estimates of the model's output statistics can be calculated. Three of these techniques use polynomial chaos (PC) expansions to construct the model proxy, but they differ in their approach to determining the expansions' coefficients; the fourth technique uses Gaussian Process Regression (GPR). An integral plume model for simulating the Deepwater Horizon oil-gas blowout provides examples for illustrating the different techniques. A Monte Carlo ensemble of 50,000 model simulations is used for gauging the performance of the different proxies. The examples illustrate how regression-based techniques can outperform projection-based techniques when the model output is noisy. They also demonstrate that robust uncertainty analysis can be performed at a fraction of the cost of the Monte Carlo calculation.

KW - Gaussian processes

KW - integral plume model

KW - polynomial chaos

KW - uncertainty quantification

UR - http://www.scopus.com/inward/record.url?scp=84968906590&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84968906590&partnerID=8YFLogxK

U2 - 10.1002/2015JC011366

DO - 10.1002/2015JC011366

M3 - Article

VL - 121

SP - 2789

EP - 2808

JO - Journal of Geophysical Research: Oceans

JF - Journal of Geophysical Research: Oceans

SN - 2169-9275

IS - 4

ER -