On 26 April 1999, an area of showers and strong thunderstorms developed over southwest Texas during the early morning hours. This area quickly grew into a large complex of thunderstorms, with a squall line along its leading edge. The convective system moved rapidly eastward across northern and central Texas, and by early afternoon the leading edge of the storms had moved into the Arklatex region. As dry westerly winds moved in behind the convection, the skies rapidly cleared across the western half of Texas.

I noticed something I had never seen on satellite imagery during that afternoon. If you look at the visible satellite picture of Texas from 1815 UTC, you can clearly see that the sky had cleared over the west half of the state. A new complex of thunderstorms had developed north of Corpus Christi, while the old line of thunderstorms was entering Arkansas and Louisiana. Now if you look at the infrared (IR) satellite imagery from the same time period, the sky over west Texas at first glance does not appear to be clear. There is a lighter shade of color which suggests a colder IR temperature than the surrounding area. In early afternoon this often suggests the location of lower clouds which are sensed within the infrared spectrum as being colder than the surrounding land temperatures. There appears to be a low cloud deck east of a line from Del Rio TX to east of Childress. This, however, did not match what was observed on the visible image, and the metar surface observations over west Texas reported clear skies. So what was the infrared imagery showing?

The answer is confirmed by examining the accumulated rainfall distribution map. The last figure is the 24-hour rainfall as shown by the Stage III Precipitation Processing System at the West Gulf RFC. One can see that over the south part of west Texas, there were two distinct bands of rainfall. One extended from southeast of Pecos TX to south of Midland, while the other extended from west of Ozona TX to San Angelo. If you compare the location of these rain bands as seen on Stage III with the lighter colored area on the IR satellite imagery, they match. What the IR satellite imagery is sensing is the difference in temperatures between the hotter land areas (roughly 32 °C) which received no rainfall and the cooler land area where the rain fell and evapotranspiration was likely occurring (roughly 24 °C).

The difference between the rain/no rain areas as sensed by the IR satellite imagery can be important. Dr. Ken Crawford gave a presentation on this subject back in August, 1996 at the Southern Region QPF Workshop. Dr. Crawford demonstrated case studies in Oklahoma where the difference in temperatures between areas which received rainfall and those that did not set up boundaries which led to short-term climate differences. One often does not think of these subtle temperature differences to be important, but mesoscale baroclinic zones can set up in the areas where land temperatures differ, which can lead to the development of future convection.