Variational Assimilation of XBT Data. Part 1

Julio Sheinbaum Hooke Institute for Atmospheric Research and Department of Atmospheric, Oceanic and Planetary Physics, Clarendon Laboratory, Oxford, U.K.

Search for other papers by Julio Sheinbaum in
Current site
Google Scholar
PubMed
Close
and
David L. T. Anderson Hooke Institute for Atmospheric Research and Department of Atmospheric, Oceanic and Planetary Physics, Clarendon Laboratory, Oxford, U.K.

Search for other papers by David L. T. Anderson in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

A variations method based on the adjoint equation technique is used to assimilate data in a relatively simple linear reduced gravity model of the tropical Pacific. Real XBT data are used by identifying the depth of the 16°C isotherm depth with the model layer depth. It is shown that the XBT data contain large scale information that corrects the model first guess. However, the model is not capable of fitting the data in the eastern Pacific for the whole assimilation period. Regions not seeded by the data are explicitly shown and the impact of data from different times on the initial state is also discussed.

Abstract

A variations method based on the adjoint equation technique is used to assimilate data in a relatively simple linear reduced gravity model of the tropical Pacific. Real XBT data are used by identifying the depth of the 16°C isotherm depth with the model layer depth. It is shown that the XBT data contain large scale information that corrects the model first guess. However, the model is not capable of fitting the data in the eastern Pacific for the whole assimilation period. Regions not seeded by the data are explicitly shown and the impact of data from different times on the initial state is also discussed.

Save