iTOUGH2 Course

iTOUGH2 Inverse Modeling Training Course for Scientists and Engineers Short Course

A iTOUGH2 Inverse Modeling Training Course for Scientists and Engineers Short Course will be held at Lawrence Berkeley National Laboratory, Berkeley, California. Note that space is limited and registration is on a first-come, first-served basis.

Registration Cost: General Attendee $1,250; Student $600

COURSE AGENDA

NOTES:

• Information provided below is meant to give you a sense of what to expect during this course (but is subject to change).
• This course is scheduled for October 3-5, 2018.
• Space is limited and registration is on a first-come, first-served basis.

Course Objectives

The training course provides the participants with the theoretical background and practical knowledge needed to solve inverse problems in geosciences and engineering. The lectures cover all aspects related to automatic model calibration using nonlinear parameter estimation methods, including sensitivity, data-worth, error and uncertainty propagation analyses. The lectures are complemented with computer exercises using the iTOUGH2 simulation-optimization code.

Course Outline

The introductory lectures will discuss the process of parameter estimation by automatic calibration of numerical models used in geosciences. Parameterization of the forward model, selection of calibration data and their incorporation into an objective function are essential issues, as they determine whether a well-posed inverse problem can be formulated. Multiple optimization algorithms suitable to minimize the objective function will be presented. The need and usefulness of performing formal sensitivity, residual, error, and uncertainty analyses will be emphasized. In addition to the lectures, which provide the theoretical background, computer exercises using the iTOUGH2 simulation-optimization code will provide hands-on experience in solving practical inverse problems. Moreover, the course participants may be asked to work on a small project (either individually or in small groups), in which they formulate a tractable optimization problem and solve it using iTOUGH2. The combination of lectures, computer exercises, and project work will provide the attendees with solid knowledge in inverse modeling and initial experience in the use of iTOUGH2 for solving a variety of parameter estimation and optimization problems

Prerequisites

Understanding and experience in numerical simulation and modeling; good knowledge of the TOUGH2 nonisothermal multiphase flow simulator or completion of the TOUGH Beginner’s Course.

Material to be Distributed during the Course

Participants are required to bring their own laptop. Course attendees will be provided with hard copies of the lecture notes. Electronic copies of the lecture notes, all relevant manuals, as well as input files used during the course will be made available. The software will be made available for installation on the participants’ laptop computers used during the Short Course. Course participants need a valid license to use the software after the course. For more information, please visit our TOUGH website and/or Berkeley Lab Software Center.