What's the best way to handle missing values due to a qualityControl value of "Z"

[oester]

When pulling current observations for a station, I run into values (wind speed being the most common) with a qualityControl value of "Z". For instance the wind value was there the previous fetch, but now it's missing. I'm wondering how best to handle this when I'm displaying conditions for a given location. My current thinking is that if a value is missing due to "Z" status just display the previous value, and then update it when the new value is present.

[rr0ss0rr]

What I've done in the past for missing values is to retrieve the values from the METAR record from aviationweather.gov. Of course, this works if the location is close to an airport ;-). What I do now is routinely use the values in the METAR record as I find it consistently offers more info with a faster API turnaround time

[oester]

Thanks. That does look like a good solution and things like windchill can be calculated from values give. Faster and more reliable than the NWS! I'll use METAR and all back to NWS if it fails.

[rr0ss0rr]

Almost guaranteed that the dew point will be reported on the Metar (and it's listed twice). Unfortunately, all the interesting stuff comes after the RMK identifier which the API does not decode. the Federal Meteorological Handbook No. 1 documents most of the RMK fields. There's the python-metar decoder repo on GitHub and I wrote a decoder in bash. Let me know if you have any questions.

[oester]

Where can I find that python script? I think that the base METAR information is probably good enough, since I can calculate the missing values that NWS has.

[rr0ss0rr]

Python-metar

[oester]

Yes, Remarks looks valuable. Is there a way to pull the raw METAR data for a station via URL so it can be decode by python-metar?

[rr0ss0rr]

sure .. Replace KROC with your station id "https://aviationweather.gov/api/data/metar?ids=KROC&hours=0&sep=true"