Some UCAR models may be broken

[nwagenbrenner]

It maybe just be NDFD, but we need to double check all UCAR models are working properly.

[masonwillman]

I tested the following curl request, and this was the response:
Request:
https://thredds.ucar.edu/thredds/ncss/grid/grib/NCEP/NDFD/NWS/CONUS/NOAAPORT/Best?north=48.0757&west=-115.186&east=-112.881&south=45.7973&time_start=present&time_duration=PT6H&accept=netcdf3&var=Total_cloud_cover_entire_atmosphere_single_layer_layer,Wind_direction_from_which_blowing_height_above_ground,Wind_speed_height_above_ground
Response:
Throwable exception handled : java.lang.NullPointerException

Additionally, when I attempted to access the data on the THREDDS data center, I recieved the following message:
FileNotFound: No such file or directory

I have reached out to THREDDS support to see if they still host the NDFD model. If not, I have asked if they know of other hosts that we could leverage. Will update with their response.

[masonwillman]

Received a response:

Greetings Mason!

Thank you for your report. It looks like there was a somewhat silent change to the data feed upstream for NOAAPORT NDFD grids, which is why they disappeared from our server. We're working to get things restored and data should start flowing again today.

Cheers,
Sean

[masonwillman]

Further responses:

Quick update. I've been able to restore the feed from NOAAPORT, but some of the GRIB messages that are getting delivered are invalid. We're trying to get that sorted and, once addressed, we'll be able to get this served out once again.

In the meantime, are they data you are looking for in the NDFD collection that we have from CONDUIT (direct LDM feed from NCEP as opposed to the satellite broadcast)?

https://thredds.ucar.edu/thredds/catalog/grib/NCEP/NDFD/NWS/CONUS/CONDUIT/catalog.html

I believe this collection will generally have more fields available than what comes across NOAAPORT and may serve your needs even better than the NOAAPORT version. (Note the data on thredds.ucar.edu are a little out of date at the moment, but should be updated shortly).

I do not anticipate that this will be sorted out by tomorrow, so, regrettably, I would plan on disabling NDFD for now. I will keep you updated as we make progress with NOAA.

I do think it is worth checking out the CONDUIT collection, as I believe this feed is also what powers the NOAA NDFD FTP/HTTP sites. NOAAPORT should just be a subset of the NDFD, whereas CONDUIT should contain everything. To give you an idea of what is available, you can interactively explore/display the data from the CONDUIT collection on our TDS using the web-based Godiva3 client:

https://thredds.ucar.edu/thredds/Godiva.html?server=https%3A%2F%2Fthredds.ucar.edu%2Fthredds%2Fwms%2Fgrib%2FNCEP%2FNDFD%2FNWS%2FCONUS%2FCONDUIT%2FNDFD_NWS_CONUS_conduit_2p5km_20260305_1600.grib2

[masonwillman]

CONDUIT could be a potential alternative. If we navigate to a .grib file and select "NetcdfSubset," you can see the request url taking in coordinates to request a subset of the data.

Example:
https://thredds.ucar.edu/thredds/ncss/grid/grib/NCEP/NDFD/NWS/CONUS/CONDUIT/NDFD_NWS_CONUS_conduit_2p5km_20260312_1600.grib2?north=52.818&west=-130.123&east=-60.866&south=20.179&horizStride=1&time_start=2026-03-12T17:00:00Z&time_end=2026-03-15T12:00:00Z&&&accept=netcdf3