Greetings,
The wild goose chase I am on today is to audit my processing of NGRID
products and see why some LDM productIDs use numbers and not character
descriptors for some grids. So lets start by auditing the NGRID products
for 12 UTC on 14 September 2016.
As a refresher, here's an example productID
LEMF98 KWNH 141200 !grib2/ncep/NCEP_QPF/#001/201609141200F030/APCP06/0 -
NONE
So I am interested in using the grib2 translated string of
'grib2/ncep/NCEP_QPF/#001'
Here's a listing of unique portions of the expanded LDM productIDs:
grib2/161/161/#255
grib2/ncep/118/#255
grib2/ncep/15/#255
grib2/ncep/17/#255
grib2/ncep/221/#001
grib2/ncep/AWC_CIP/#130
grib2/ncep/CFS/#255
grib2/ncep/DGEX_115/#186
grib2/ncep/ESTOFS/#255
grib2/ncep/FORECASTER/#255
grib2/ncep/GEFS/#003
grib2/ncep/GFS/#003
grib2/ncep/GFS/#160
grib2/ncep/GFS/#161
grib2/ncep/GFS/#197
grib2/ncep/GFS/#213
grib2/ncep/GFS/#215
grib2/ncep/GFS/#254
grib2/ncep/GFS/#255
grib2/ncep/GMGWM/#255
grib2/ncep/HRRR/#255
grib2/ncep/LAMP/#197
grib2/ncep/LAMP/#255
grib2/ncep/NAM_84/#218
grib2/ncep/NAM_84/#242
grib2/ncep/NCEP_QPF/#001
grib2/ncep/NDFD/#255
grib2/ncep/NMM_89/#255
grib2/ncep/RTMA/#197
grib2/ncep/RTMA/#255
grib2/ncep/RUC2/#130
grib2/ncep/SSIGFS/#003
grib2/ncep/SSIGFS/#160
grib2/ncep/SSIGFS/#161
grib2/ncep/SSIGFS/#213
grib2/ncep/SSIGFS/#215
grib2/ncep/SSIGFS/#254
grib2/ncep/SSIGFS/#255
grib2/ncep/WRF_EM/#255
grib2/ncep/WRF_NMM/#255
So we have five entries of interest:
grib2/161/161/#255
grib2/ncep/118/#255
grib2/ncep/15/#255
grib2/ncep/17/#255
grib2/ncep/221/#001
but lets backup, how are these values set in LDM... I don't have a
noaaport dish to test things to be sure, but I think it is here:
https://github.com/Unidata/LDM/blob/master/noaaport/grib2name.c
So the s_pds_center() and s_pds_model() appear to be important :) So our
journey leads here:
https://github.com/Unidata/LDM/blob/master/grib2/gribid.c#L30
So '161' is not defined there. Ah yes, the pain of grib table
authorities. So perhaps the listing is here:
http://www.nco.ncep.noaa.gov/pmb/docs/on388/table0.html
161 is 'US NOAA Office of Oceanic and Atmospheric Research', is there a
convention on what that should be abbreviated to for the purposes of
gribid.c tables? So now onto the unknown models
Are those here? http://www.nco.ncep.noaa.gov/pmb/docs/on388/tablea.html
161 'Reserved'
118 URMA (UnRestricted Mesoscale Analysis)
15 Nearshore Wave Prediction System (NWPS)
17 Extra-tropical Storm Surge Pacific Domain
221 Reserved for WPC products
Do folks have thoughts on this situation? Perhaps Unidata has some
procedure for when and how they update LDM to add these additional IDs
into the lookup tables?
daryl
--
/**
* daryl herzmann
* Systems Analyst III -- Iowa Environmental Mesonet
* https://mesonet.agron.iastate.edu
*/
