Bad Soundings

[Bob Hartig]

At the COD severe weather symposium, Roger Edwards commented about bad sounding data. It sounds like one Rawinsonde station in particular gives consistently bad data, but Roger never specified its location and I never asked him. Does anyone here know?

And, what are some common kinds of bad sounding data? I'm using skew-T's more and more, but I've still got a few things to learn about interpreting them, and I'm not at a point yet where I can look at one and think, "Hmm...faulty read." So any tips will be appreciated, particularly visual examples.

 

[rdale]

Probably related to the new sounding replacement system.

http://stormtrack.org/forum/showthread.php?t=18333

http://stormtrack.org/forum/showthread.php?t=13572

http://stormtrack.org/forum/showthread.php?t=13051

http://stormtrack.org/forum/showthread.php?t=13246

 

[Bob Hartig]

Thanks for the links, Rob. Looks like I've got some reading to do, but just the little I've glimpsed so far suggests that the problem is pretty widespread.

 

[rdale]

Kudos to Bob Maddox for really working on this project. Unfortunately the official NWS line is "there is nothing to see here, move along."

 

[Bob Hartig]

Good grief! I just finished reading through the links you supplied, Rob, including those embedded in the four threads, and I am absolutely floored. If that's the kind of data that is being fed into the forecast models, how on earth can we trust anything other than surface obs? Models, schmodels. This is scary!

The Tucson fiasco in particular just blows me away. Comparing Bob Maddox's description with site guidelines makes me wonder what on earth NWS officials were thinking, since the violation is so flagrant.

The most recent of the four threads looks to have been about a year ago, three years after implementation of the "new and improved" Sippicon sounding devices. Has there been any improvement at all, both in troubleshooting the technology and in the outrageously lame-ass non-communication of the NWS?

 

[rdale]

I guess since it doesn't happen a majority of the time is a good thing, so I wouldn't throw them out, but it certainly doesn't look good.

There is a data quality page updated monthly at http://www.nws.noaa.gov/ops2/ops24/rrs.htm