An investigation into the potential of miller's technique for forecasting large hail
A sample of 16 cases in the UK has been used to test the accuracy of Miller's technique for forecasting large hail using Mesoscale Model data. Miller's method dates back to 1972 and is normally presented as a nomogram for use by forecasters. However, for this study the procedure was coded up as a FORTRAN subroutine. Out of the sixteen cases, eight were when large hail fell somewhere in the UK and eight were when no large hail was reported. The hit rate in the experiments was 62.5% with a false alarm rate of 37.5%. However, the key result was that in the five cases when large hail was both forecast and observed, the Mesoscale Model was able to pinpoint single occurrences of large hail very accurately. The model was also able to provide a useful indication of size, up to 9 hours ahead in one case. The experiment results are extremely encouraging and demonstrate that the Mesoscale Model is capable of being used to accurately diagnose severe weather on short. time-scales. Consequently it has been recommended that this technique be incorporated into Nimrod and be used as a basis for providing an operational hail size output diagnostic from the Met Office Mesoscale Model.
Accès à la notice sur le site du portail documentaire de Météo-France