found and fixed bug in r.out.gdal export of netcdf files with no missing data. The missing data value was set to 0 instead of the 255 that was specified. This resulted in all 0s to be removed from these day-tiles. Fixed by deleting and resetting the attribute _FillValue using ncatted.
Updated to use new swtif and remove additional QC that identified strips introduced by old swtif. Some tiling artifacts still present.
Adding script to upload covariate data to Pleiades
updated swtif error checking to use r.neighbor instead of r.mapcalc. Submitted job for northern block of tiles
Add first version of script to develop global modis MODLAND grid
added function to flag bad pixels introduced by bug in HEG software
Started process of adding sensor zenith filter to MOD35 Processing
Separated daily products into a 'day' and 'night' cloudiness based on day flag
Move file outputs to /nobackupp1 instead of lou due to quota increase. First full 2009 summary
Updated swath-grid section of MOD35 processing to add buffer and (hopefully) reduce edge artifacts in tiles
updated MOD35 processing script to work in separate temporary directories and transfer daily cloud data to lou for archiving in preparation for global processing (which cannot be done in personal directory). Also set up MOD35_Climatology script to be submitted to LDAN queue to run on lou rather than Pleiades. However environment still not set up correctly and script fails when reading netcdf files. Andrew is working on it.
Fixed the 'missing stripes' problem in the climatology script. Source of problem was related to the _FillValue attribute being incorrectly set to a short integer when the data was saved as byte. Changing both the attribute and the data to type 'byte' appears to have fixed the problem.
Removed some unsed code from MOD35_climatology. Still having problem of missing data stripes in February means.
Updated MOD35_Climatology to use percentages instead of classes, but there are too many zeros in result
Simplifications to Climatology Script
