You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
With the addition of the geo-stationary sst's, it take ~12mn to process the satellite obs for 1 cycle. This needs to be parallelized.
Executables of interest:
Sets up the ocean obs prep task to run the NetCDF to IODA converter
executable in parallel by obs space, node permitting. Also tries to
detect and notify of failures.
Tested with ctests on Hera, including pointing `DMPDIR` at
`/scratch1/NCEPDEV/stmp4/Shastri.Paturi/forAndrew/`. Runtime was reduced
only from 16 minutes to 15 minutes, but that is probably because only
rads and metop b GHRSST are processed, with the latter being the
bottleneck.
Partially addresses #886
I'm getting timing around 230 sec @AndrewEichmann-NOAA , that's with 2 AVHRR and 2 VIIRS but no geostationary and a stride of 15 for binning. Were you able to run a test on your side?
With the addition of the geo-stationary sst's, it take ~12mn to process the satellite obs for 1 cycle. This needs to be parallelized.
Executables of interest:
GDASApp/scripts/exglobal_prep_ocean_obs.py
Line 85 in 0ceff32
The text was updated successfully, but these errors were encountered: