Custom Query (964 matches)

I added couple of scripts for computing the 4D-Var error covariance standard deviations which are used to convert modeled correlations into covariances.

• 4dvar/d_std.m: driver template to compute and write 4D-Var standard deviations.

• 4dvar/d_std_unbalanced.m: driver template to compute and write 4D-Var unbalanced standard deviations.

I also updated c_grid.m, c_observations.m, and c_std.m.

Several new scripts were added to the matlab/4dvar repository:

• 4dvar/c_std_bry.m: creates 4D-Var open boundary conditions standard deviations NetCDF file.

• 4dvar/c_std_frc.m: creates 4D-Var surface forcing standard deviations NetCDF file.

• 4dvar/d_sst_obs.m: driver template to extract SST observations from satellite data. It creates and writes observation file. Then, it computes super observations and creates and writes super observations NetCDF file.

• 4dvar/d_std_frc.m: driver template to compute and write 4D-Var standard deviations for surface forcing.

• 4dvar/load_sst_pfeg.m: extracts satellite sea surface temperature for the period of interest and specified region from ROMS Grid file. The SST data is from the OpenDAP catalog maintained by NOAA PFEG Coastwatch in California. The resolution is 0.1 degree global 5-day average composite.

I also renamed load_ssh_data.m to a more appropriated load_ssh_aviso.m. I made minor documentation corrections to several scripts.

Many thanks to John Wilkin for providing a template for load_sst_pfeg.m.

I updated several scripts in the matlab/4dvar and matlab/utility directories of the repository to enhance their capabilities. This stuff is complicated and the devil is in the details.

We need for these scripts to be flexible. I assume that more fine tunning is required when the users start using them next week at the 4D-Var workshop.

I added the grid_Lm_Mm_N global attribute to the 4D-Var observation metadata. This is needed in quality control and binning into super observations.