Data Assimilation Meetings at Reading
Date | Meeting type | Speakers |
17 October 2012 |
DARC tutorial room change: Met 1L43 |
Amos Lawless (UoR) How to code an adjoint model: A tutorial Tangent linear and adjoint models are widely used in environmental prediction schemes. They appear in data assimilation schemes based on the variational method (e.g. 4D-Var, 4D-PSAS), in forecast sensitivity studies and in singular vector calculations for ensemble forecasting. In this lecture I will explain how to derive these models starting from the source code of a nonlinear model. Robust methods for testing the correctness of the models will also be explained and illustrated. This tutorial is aimed at those who have never coded an adjoint model but would like to do so. |