Data Assimilation Meetings at Reading
Date | Meeting type | Speakers |
21 January 2015 1L61 | Internal speaker |
Peter Jan van Leeuwen (University of Reading)
Representation errors in linear and nonlinear data assimilation Representation errors are related to the different representation of the truth in model and observations. As such, they arise in the likelihood in the Bayesian formulation of data assimilation, and not in the prior. We will discuss the assumptions needed in the usual procedure to add a representation error covariance to the error covariance of the observations. Specifically, it is shown that when several sub-grid observations are present their mean still has a representation error, so so-called ’superobbing’ does not resolve the issue. Connection is made to the off-line or on-line retrieval problem, providing a new simple proof of the equivalence of assimilating linear retrievals and original observations. Furthermore, it is shown how nonlinear retrievals can be assimilated without loss of information. Finally we discuss how errors in the observation operator model can be treated consistently in the Bayesian framework, connecting to previous work in this area. |