to prepare a database of ocean observations
TEST data in [/test_data/profile] comes from:
wget http://www.usgodae.org/pub/outgoing/fnmoc/data/ocn/profile/2000/2000061800.profile.Z uncompress 2000061800.profile.Z
SOURCE CODE to decode and CHECK decoder: [test_data/docs/] http://www.usgodae.org/ftp/outgoing/fnmoc/data/ocn/docs/ wget http://www.usgodae.org/ftp/outgoing/fnmoc/data/ocn/docs/report.profile mv report.profile report.profile_2000061800 ** reproduced above, see decoder/ocn_obs.x; additional columns do NOT MEAN NON-ZERO diff
Data availability (note: ARGO starts sometime in 2000), so all good from USGODAE!
Before 1998, there's is NO data from the USGODAE- fnmoc. For profiles, I suggest one gets the best possible (QC'ed and bias corrected) obs from EN4 https://www.metoffice.gov.uk/hadobs/en4/download-en4-2-1.html <-- data URL https://www.metoffice.gov.uk/hadobs/en4/en4-0-2-profile-file-format.html <-- format info which goes back to 1900s.
For ships, buoys, etc, Please go to ICOADS: http://icoads.noaa.gov/products.html most of the surface obs have been processed into IQUAM, so that could be a easier/quicker way to get the SURFACE obs.
option to python plot- needs to READ netcdf output: horizontal spatial map of locations, vertical profile map of T (and S)-- total of 3 plots, like from Coriolis