This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.
Mark, The GARP_PATH environmental variable is typically set by sourcing the Gemenviron file as: setenv GARP_PATH $GARPHOME/config Garp uses the variable to look for the Garp_defaults file in the above directory. If you obtain case studies, or have archived data with different file name than your Garp_defaults expects with its keys, then you can create an alternate Garp_defaults file and place it in another location and reset your GARP_PATH variable before launching Garp. I suspect case study data sets may need different keys than Unidata real-time data sets- for instance they might have MM5, NOGAPS, or other models, mesonets or different surface networks etc, so the ability to use an alternate GARP_PATH variable is definitely needed. Steve Chiswell Unidata User Support >From: Mark Tucker <address@hidden> >Organization: . >Keywords: 199907201946.NAA00402 >On Fri, 16 Jul 1999, Unidata Support wrote: > >> >> Check for differences in "NIDS" vs "nids" as directory names. Note that you >> have to restart Garp after changing the defaults file. > >The path was lower-case. I should have seen this earlier. > >> >> >> > >> >2. Model data. This problem is similar to the NIDS issue above. >> >Currently we can only see AVN and ECMWF. No changes to our data >> >ingestion have been made. >> >> When you say you are updating your lab, are you loading a new version of >> the Gempak distribution? I have changed some of the model keys in >> the pl11 distribution since NOAAport has several ETA grids, AVN, MRF etc. >> >> Peviously the model key would be "eta", now it is "eta_grid211" with >> similar changes in the pqact.conf entry for dcgrib naming the >> files with YYYYMMDDHH_eta_grid@@@.gem. >> >> If this isn't it, then it would help to see the output of: >> ls $HDS >> and your Garp_defaults. >> > >This appears to be the issue. We still have an older version of gempak >doing our decoding so the model keys are still naming with the >YYMMDDHH_eta.gem format. I can (and probably will) upgrade gempak on our >server along with the necessary changes to the pqact.conf file. This does >leave one significant issue: >How can we use archived data, from case studies and local archives, with >the current version of gempak? Can the model key be temporarily changed >on the workstations? If so, how? > >Mark Tucker >Information Technology >Lyndon State College >address@hidden > >