Identification and early warning of tornadic storms by geostationary satellite-inferred cloud microstructure

The prime reason for convective clouds becoming severe is the fast conversion of static instability into kinetic energy of fast rising cloud elements. The fast vertical motions profoundly affect cloud microstructure at the mid and high elevations, because the cloud drops have little time to grow by coalescence and to be accreted on ice hydrometeors. This has been already previously documented and recognized as a "severe storm microphysical vertical profile" using polar-orbiting satellites. However, practical application of this finding can be realized only with geostationary satellites because of their excellent temporal coverage.

Retrieving the dependence of cloud top indicated effective radius (re) as a function of cloud top temperature (T) can provide the T-re relations and these can be classified into “severe" and “non-severe” signatures.

Analyses of documented cases of large hail and tornadic storms will be presented for GOES multi-spectral satellite imagery. It is shown that the "severe storm" microphysical signature can be detected as soon as the first cloud elements in the pre-storm environment reach the -40C level. This occurs typically at least one hour before the actual occurrence of the severe storm phenomenon on the ground.

The GOES-retrieved T-re relations for 86 individual Cb clusters, 34 of which produced a total of 78 tornadoes, were analyzed for their tornado predictive skill. The results show that:

1. The T-re relations obtained from the GOES still provide similar qualitative severe storm microphysical signatures as the AVHRR does (which is reported in a separate presentation), in spite of the poorer spatial resolution.

2. A tornado predictor that is based on a logistic regression with the parameterized T-re relation is highly correlated with the occurrence of the tornados.

3. The skill of the GOES based predictor exceeds that of conventional sounding based predictors, and is more focused in time and space, giving advance information when and where the tornado is about to occur..

The new methodology can fill the gap between a watch and a warning. The encouraging results justify the continued development of the methodology into an operational system. It also provides a new incentive for the improved spatial and temporal resolution of future geostationary satellites, because the accuracy of the GOES predictors appears to be limited by the poor spatial resolution in the infra-red channels of existing GOES satellites.