dpaulat / supercell-wx

Supercell Wx is a free, open source advanced weather radar viewer.

https://supercell-wx.rtfd.io

MIT License

85 stars 14 forks source link

Updated radar locations and added elevation data in radar_sites.json using NOAA's HOMR #233

Closed AdenKoperczak closed 3 days ago

AdenKoperczak commented 1 week ago

For issue "Fix Radar Site Locations #132 " Updated radar locations and added elevation data in radar_sites.json using NOAA's HOMR. I wrote a Python script to make these changes. It pulls from the API directly. I would be willing to add it to the repository. Elevation data is elevation of the ground in feet. KVBG, KHDC, and KLIX were labeled as AWOS (Automatic Weather Observing System) or UPPERAIR. I did pull the data for them, but KVBG and KHDC did not include ground elevation data.

AdenKoperczak commented 1 week ago

KVBG

Is this possibly supposed to be KVBX not KVBG? I am not getting any radar on KVBG, but I got radar on KVBX when I manually added it . Both are named Vandenberg AFB n the HOMR data, and KVBX is labeled as NEXRAD.

KHDC and KLIX

It looks like KLIX was moved to Hammond. According to the HOMR data, it looks like it was to be renamed to KHDC. Unfortunately, it does not look like the location for KHDC has been uploaded to the HOMR datapase. The current location they have on another document is in the center of the Hammond Municipal Airport (based on the map), and is likely somewhat off from the actual location, but I do not have a better source.

links

AdenKoperczak commented 1 week ago

Possible better location source for KHDC. https://www.weather.gov/media/notification/pdf_2023_24/scn24-11_khdc_nexrad_level_iii_product_dissemination.pdf

dpaulat commented 1 week ago

KVBG

KVBX is correct for the current site per NOAA, and cross-checked with RadarScope: https://radar2pub.ncep.noaa.gov/ I see references noting KVBG, but I'm not sure what the difference is. I don't see a KVBG in the AWS level 2 archive.

I doubt this radar site has ever worked... and nobody reported it until now!

KHDC and KLIX

KLIX doesn't match the HOMR location. It looks like it was updated to the runway location instead.

KHDC was added as part of this issue: https://github.com/dpaulat/supercell-wx/issues/86 It looks like the SCN PDF you posted has a good location, but I don't know of any satellite imagery that's been updated with the new dome to confirm. I don't think either the current location or the new location you have is quite right - they are too close to the runway.

Misc

I spot checked a couple other sites and they looked OK. Aside from what's noted, do you expect them to all be HOMR coordinates?

AdenKoperczak commented 1 week ago

Everything else looks good to me. I did run through all the stations briefly on Google Maps and did not notice anything else that looked off, but I was moving quickly. I will switch KVBG to KVBX, grab the old coordinates for KLIX (It is labeled as inactive, so my script grabbed the runway coordinates which are active), update KHDC from the document and update my pull request.

My only other question is if you would like me to clean up my script and add it to the repository for future use, and if so, where do you want it? (Python3 using requests library).

dpaulat commented 1 week ago

Great, thank you! I think the tool could probably go under scwx-qt/tools/.

AdenKoperczak commented 1 week ago

On KLSX and general location precision. The HOMR data reports its precision to be "DDMMSSss", which I interpreted as seconds to the hundredth place. That being said, the coordinates for many of the stations (including KLSX) lie on a round second, (even when converted from the decimal form they provide). They may actually only reported down to seconds. Furthermore, the difference between KLSX's old and HOMR coordinates are less than a second. Although some stations are more precise, I have not found more precise data on HOMR for those that are only to seconds, so maybe a different source should be found if you want more precision. I will work on updating the test for this pull request, then your other suggestions when I have time this week.

dpaulat commented 3 days ago

It looks like the locations have improved for about 3/4 of the sites. It does look like some WSR-88D sites do only have second precision, but they are still pretty close. Unfortunately the second/subsecond precision doesn't matter much when the location was not completely accurate to begin with.