Sensitivity Analysis and Parameter Tuning of a Sea Ice Model
by
Paul Hovland
Argonne National Laboratory
Coauthors: Jong Kim

The values used for many of the parameters in climate models are often not known with any great precision. We are therefore exploring the possibility of using sensitivities computed via automatic differentiation to solve the inverse problem of identifying parameter values for which numerical simulation provides results as close as possible to observational data. Our candidate model is a sea ice model from the University of Colorado. We are using ADIFOR to differentiate the model and the L-BFGS-B algorithm to perform the optimization. Preliminary results, using simulated observational data, are promising. Current research focuses on applying the same method using real data.

Extended abstract for Sensitivity Analysis of Sea Ice Model (postscript)

Date received: February 11, 2000

Copyright © 2000 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cads-76.