3.1.8.2. Ocean Model Assimilation

During the investigation period, we have conducted studies on data assimilation techniques using eddy-resolving OGCMs. The Princeton Modular Ocean Model (MOM) has been implemented to run on the UT Cray computer and on a CSR RS-6000 workstation. Although MOM is a rigid-lid model, which does not conserve mass, we are using it as a tool to study data assimilation techniques. In this study, the model has been initialized by using climatological data. The initial conditions for the model were zero currents and the climatological temperature and salinity fields developed by Levitus. The model was also forced with climatological wind stress developed by Hellerman. The implementation used a 1° horizontal resolution, 15 vertical levels, real topography, with a second order horizontal mixing scheme with constant eddy coefficients for momentum and for heat transport and diffusion. Sea surface climatological temperature was used as a boundary condition. The three-year model initiation run was used to obtain a climatological distribution of ocean currents. As a first step to satellite data assimilation in the model, ERS-1 scatterometer vector wind measurements, and the altimeter sea surface topography measurements available from the TOPEX/ Poseidon altimeter were used. We first used the ERS-1 computed wind stresses as boundary conditions. The data were filtered and subjectively analyzed. Using the 3-year spun up model, the ocean model was then forced with monthly averaged scatterometer wind fields (October 1992-March 1993). A comparison of the model output with the in situ wind measurements recorded in the Western Tropical Pacific for the same period indicated good agreement. Since MOM is not a free-surface model, we used the sea surface heights observed from TOPEX/Poseidon altimetry in the form of temperature fields to update the solution at various time steps. The ECMWF wind fields are now being prepared for use in the model and to perform further comparison studies. We expect to conduct further data assimilation studies using an improved version of MOM (MOM2) which is a free-surface model.

3.1.9. Specific Research Plans Before First EOS Launch

At present, we are actively involved in analyzing a wide range of existing satellite data, which include pre-EOS sensor data. These include altimeter data from Seasat, Geosat, TOPEX/Poseidon, ERS-1, ERS-2, and other pre-EOS sensor data including Lageos-class satellite laser ranging data, and AVHRR thermal images from the NOAA polar-orbiting satellites. We will collect and analyze data from SeaWifs when it is launched. We are also analyzing an existing set of tracking data, both radiometric from the Doris and GPS systems, and satellite laser ranging data collected over other cannonball geodetic satellites. We are involved in analyzing data from tide gauge measurements and a wide range of other data sets collected under the WOCE and TOGA programmatic activities. Finally, we are analyzing a wide range of meteorological and geophysical data and output from GCMs, including NMC (reanalysis), ECMWF, and GEOS for Atmospheric Angular Momentum and other studies.

Specific research activities during the next two-year period will be focused on the following: