An active spring weather pattern was in place across the southwestern and central U.S. on the morning of 22 April 2001, with a deep upper trough and closed low situated near the Four Corners region and a broad belt of southerly low level flow (25-30 m/s at 850mb) from south Texas into the Central Plains. Forecasters at the Storm Prediction Center (SPC) noted that wind shear and helicity profiles over western and central Oklahoma were favorable for the development of severe thunderstorms, including isolated supercells, but they expected linear structures (i.e., a squall line) to be the dominant convective mode. In particular, they noted a pronounced elevated mixed layer over the Southern Plains at the 1200 UTC sounding time, with 700 mb temperatures approaching 15oC, so they anticipated that associated convective inhibition (CIN) would suppress thunderstorm development until a cold front swept through the area late in the day.

However, a special sounding launched at OUN (Norman, OK) at 1800 UTC caused some concern. This sounding revealed a dramatic erosion of the capping inversion, with temperatures dropping about 8oC near the base of the elevated mixed layer and CIN falling from 508 J kg-1 to just 9 J kg-1. Forecasters were compelled to re-evaluate the potential for isolated supercells ahead of the main convective line.

What physical processes could have caused this rapid erosion of the “cap”? Similar changes are observed several times a year by SPC forecasters, yet it remains characteristically difficult to provide more than a “hand-waving” explanation for this phenomenon. For example, cap erosion is often attributed to vertical motion and associated adiabatic cooling. But what causes the vertical motion? What about the role of horizontal advection? Can passing altocumulus castellanus clouds rain into dry air aloft and evaporatively cool the dry air in the elevated mixed layer? In most cases one can only speculate about the answer to these questions.

With this particular event however, much of this information can be gleaned from a numerical model. The EtaKF, an experimental version of the Eta model that is routinely provided to SPC forecasters, reproduced the OUN 1800 UTC sounding remarkably well, inspiring confidence that the model faithfully simulated the mesoscale processes associated with the cap erosion process. In this study, these processes are identified and quantified through diagnostic analyses of model output, including the extraction of temperature tendencies associated with all physical processes from the 6 h model integration.

The magnitudes of individual tendency terms have been compared and work is underway to unravel the physical processes associated with the cap erosion process. The tendencies show that evaporative cooling, associated with precipitation from a weak elevated disturbance, made the largest negative contribution to net cooling near the base of the elevated mixed layer in the model. Yet, a second model run with evaporative cooling “turned off” simulated the cap erosion process equally well. This result will be explained and the operative physical processes will be revealed at the conference. The results will be discussed in the context of improving the understanding and operational prediction of the rapid reduction of CIN.