Radiosonde 3D H(x) Evaluation

[CoryMartin-NOAA]

This issue will track the progress of evaluation of radiosonde H(x) from UFO relative to GSI for air/virtual temperature, specific humidity, surface pressure, and eastward/northward wind for UFO acceptance activities.

[CoryMartin-NOAA]

An initial look of H(x) differences for a cycle:

"Eyeballing it" looks like the 1-sigma is somewhere around 0.5m/s or 0.5K throughout the atmosphere, which seems to be too large in my opinion. This is ~an order of magnitude larger than what we are seeing with AMSU-A (see #31) where for high peaking channels differences are more like ~0.05K for 1-sigma.

[CoryMartin-NOAA]

"Eyeballing it" may be misleading. Here's a quick histogram of virtual temperature:

There are differences as large as -8 or 6K, but the mean is 0.014K and the standard deviation is 0.18K. So the differences in Tv H(x) are consistent with the differences in T GeoVaLs shown in #35