107 This commit is the final version 1 of soda2roms. This setup works correctly with ROMS so if you have to revert to an existing version this is the one to choose.
104 This is an initial version (not for full model use yet) of a script that creates a Sea Surface Salinity file from NCAR to your ROMS grid. This is just for use if you relax/nudge SSS to climatology. This routine uses the griddata module from Jeff Withaker to do an irregular 2D interpolation of the SSS to the ROMS grid. The reason for this use instead of interp was because the NCAR SSS data land matrix is too crude and leaves areas along the coast not interpolated when using interp. The griddata routine fixes that problem, but then again griddata need to be installed.
99 Important commit that solves a number of issues. First of all, all result files are now completely CF compliant. Second, a subset of the input data can now be chosen using max and min of longitude and latitude coordinates. This makes it much faster to read input data that may be global such as SODA. The solution required me to write a new module called IOsubset.py. The module contains some handy functions for getting the indices of an input grid and to combine a grid that runs from 0-360 degrees longitude into a grid that runs from -180 to 180 instead in order of making the extraction of data for the North Atlantic continuos.
98 A new module that makes the extraction of a subset of the input data more robust and generic. This modules also considers wether the domain of interest crosses the 0 degree longitude. Since SODA data are organized from 0-360, the data need to be reorganized to -180-180 for extraction and interpolation to be correct. This module solves that problem and also includes other small useful functions.
92 Added possibility of generating vstretching=2 type of s-levels. Set vstretching=2 in grd.py to use these functions instead of the default Haidvogel stretching
90 Found another bug. it turns out that the segmentation fault was caused by the fact that z_wu and z_wv wa not allocated prior to calling the routine barotropic. This is now done in soda2roms.py and the problem seems to have disappeared.
89 Found another bug. it turns out that the segmentation fault was caused by the fact that z_wu and z_wv wa not allocated prior to calling the routine barotropic. This is now done in soda2roms.py and the problem seems to have disappeared.
What changed
