Skip to main content

Skip to navigation

Departments & Programs


Quantifying Uncertainty in Subsurface Systems Workshop

Date(s):May 02, 2017
Time:8:30 AM
Location: Tresidder Memorial Union, Stanford University (Oak West Room)
459 Lagunita Dr
Stanford, CA 94305
Contact: Eiko Rutherford

Quantifying Uncertainty in Subsurface Systems

A one-day workshop, May 2, 2017, Stanford University



Uncertainty and risk are inherent to decision making for subsurface systems and hence any modeling based on which such decisions are made. This year, SCRF authors will publish a new book entitled “Quantifying Uncertainty in Subsurface Systems” with Wiley-Blackwell. This 450+ page publication covers several application areas, such as conventional and unconventional oil/gas, groundwater, contamination and geothermal energy. It uses real world cases to illustrate various strategies of decision making in the subsurface based on models of uncertainty. This workshop will feature this publication in celebration of the 30th annual meeting of the Stanford Center for Reservoir Forecasting, one day prior to the actual affiliate meeting. This one-day workshop is open to all (members, non-members, students, academics).
Quantifying Uncertainty in Subsurface Systems - Table of Contents

SCRF 2017 Workshop image


The aim is to provide an overview of real field case studies and to present the various strategies and methods developed. Attendees will be provided a first draft of the book, slated to be published in Dec 2017. More specifically, we will treat the following topics:

  • Bayesianism: most developments in science today, including uncertainty quantification are inherently Bayesian. What does this mean? How does the Bayesian application for the subsurface look like? What is prior geological uncertainty? This will be a discussion between science and philosophy.
  • Data Science for Geoscience: the book either adapts or develops modern data scientific methods to problems typical for the subsurface, such as non-Gaussianity and non-linearity. The strategies for UQ depends on methods of dimension reduction of which we will provide an overview and discussion.
  • Global sensitivity analysis: Monte Carlo & sensitivity analysis are critical to any UQ. It allows understanding the relationship between data, model and prediction variables and gain insight in the complex system of models, how they can refined or simplified.
  • Decision Science & Value of information: no practical UQ is relevant without a decision purpose. Decision science allows for scientific approaches to solving complex decision questions. We will discuss how this science applies to the subsurface including how value of information can be quantified before acquiring (possible) expensive data.


Workshop Schedule (Agenda)

Presenters: Celine Scheidt, Lewis Li, Tapan Mukerji & Jef Caers
Morning session 1

Quantifying Uncertainty in the Subsurface: overview
Decision Science & Value of information

Morning session 2

Data science for UQ in the subsurface
Global Sensitivity analysis


Afternoon session 1

UQ by Evidential learning

Afternoon session 2

Field Cases
Workshop discussion: what are best strategies/practices for UQ in the subsurface?


Registration fees

SCRF Affiliate Members / Academics: USD100
Non members: USD500
Stanford faculty, students & staff: free

Register to attend the Workshop