Early Microburst Simulations

I became interested in microburst dynamics in 1984, when Fujita's student Roger Wakimoto came to UCLA from Chicago. Roger had just published a paper on dry microbursts over the High Plains, based on data from the JAWS experiment, which indicated that the conditions favorable for dry microbursts were a deep, dry adiabatic subcloud layer, capped by a weakly conditionally unstable cloud layer. I had recently developed a 2D axisymmetric nonhydrostatic numerical model for simulating cumulus convection. We decided to see if we could simulate dry microbursts with the model.

Since the essential aspect of a dry microburst appeared to be the descent of an evaporatively cooled, negatively buoyant parcel through a deep, dry-adiabatic, subcloud layer, we set up the model to include just the subcloud layer. We specified the initial distribution of rain near the top of the subcloud layer (just below cloud base). The subcloud layer temperature and water vapor profiles were based on Roger's composite "dry microburst" sounding. In the simulation, cooling due to rain evaporation generated a negatively buoyant parcel that produced a microburst outflow.

With the "dry microburst" sounding, even very small amounts of rain water produced microburst outflows in the simulations. We next studied the sensitivity of the microburst strength (as measured by the peak outflow wind speed) to the subcloud layer lapse rate. The more stable the lapse rate, the weaker the outflow for a given initial rain water distribution. And the larger the amount of the rain, the stronger the outflow, for a given lapse rate. Srivastava's contemporary study, based on a parcel model, reached the same conclusions.

In subsequent studies, we investigated the role of ice microphysics in microbursts, the mechanism of the "wet" microburst, and the dynamics of microburst outflows.