NUCAPS: Uncorrected sounding vs obs near Rapid City

It just so happens that one of the green points on the NUCAPS Improved Latency - DB Availability product (uncorrected sounding) is very close to two locations (Rapid City airport and Ellsworth AFB) that report observations. At 1900Z (shortly before the pass time of 1913Z), both obs sites reported temps of 73F with dewpoints of 62F and 64F, respectively (first image). The 1900Z (1913Z) NUCAPS sounding (second image) that pops up when clicking on the green dot just E of these obs sites shows a way different story, with a sfc temp near 87F and a dewpoint near 54F. So, even though this NUCAPS dot was green, should it be? The data within the sounding in this instance is not good. I'm kind of souring on the benefits of using uncorrected NUCAPS soundings in an ops environment. Unfortunately, there was no corrected NUCAPS sounding for this location, so I could not accomplish a comparison. I'll try another green point near an observation site for my next blog post.

 Obs (with NUCAPS sounding "green" dot location shown just E):

Uncorrected NUCAPS sounding corresponding to the green dot:

- Thomas Bell