Assimilating altimeter sea surface height data in an operational ocean forecasting system - an historical overview
Matt Martin(1) , Adrian Hines(1) , and Mike Bell(1)
Met Office, FitzRoy Road,
EX1 3PB, Exeter,
The Forecasting Ocean Assimilation Model (FOAM) system produces operational short-range forecasts of the deep ocean every day. It was first implemented in the operational suite at the Met Office in 1997 as a global model at 1 degree horizontal resolution which assimilated in situ and satellite sea surface temperature data and in situ temperature profile data. In April 2001, a 1/3 degree resolution model of the North Atlantic and Arctic started running operationally which took its boundary conditions from the global model. This eddy permitting model was the first configuration to assimilate along-track altimeter sea surface height (SSH) data as well as the other data. More recently, various other configurations of FOAM have been run at higher resolution, including a 1/9 degree model of the North Atlantic. These configurations assimilate all the temperature and salinity profile data available over the GTS as well as the satellite altimeter data.
The along-track altimeter SSH data is assimilated in FOAM by first performing a two-dimensional analysis using the Analysis Correction scheme (an iterative approximation to Optimal Interpolation), giving SSH increments at every model grid point. For each model grid point, the water column is lifted/lowered to produce the given SSH increment such that there is no change in bottom pressure using the Cooper and Haines scheme. This paper gives an overview of the way in which altimeter data has been assimilated in the FOAM system, together with a description of the various improvements which have been made to the scheme.