Tuesday, April 19, 2011

Weather 1, Computer Model 0 (In Texas, Anyway)...

In a post earlier this evening, I commented that an experimental computer forecast model, the HRRR, was forecasting thunderstorm activity to develop over the Hill Country of southcentral Texas this evening, and reach the I-35 corridor in the Austin area by 9-10 o'clock:


Well, how did the computer forecast model perform?  Not well...


Before you get too excited, everything you see on the above radar image (just taken from the New Braunfels radar site) is radar noise, or "false echoes" (another word for ground clutter).  No thunderstorms have developed across the region this evening, thanks to a very strong capping inversion.

As far as the Lone Star State goes, it's back to the drawing board, HRRR...

I must give the computer model some credit, however, for its forecast further North.  Here is how the radar in southeast Missouri and adjancent areas looks at the same valid time as the HRRR image above:


It quite accurately forecast the thunderstorm line across Illinois, Missouri and Arkansas, advancing toward western Kentucky & Tennessee.

No comments: