engleski

Verification of global instability indices, derived from satellite data over wider Croatian area

Deep, moist convection that occurs in the 'clean air' is one of almost daily occurrences in the atmosphere of our latitudes, and elsewhere. This kind of process is common in the warmer months of the year, mostly because it is thermally conditioned and because warmer air can accept larger amounts of water vapor. Unlike e.g. the frontal convection, convection in ‘clean air’ is harder to predict. These are situations when in the morning we have 'quiet' atmosphere, but a few hours after coming to the intensive development of convective clouds. Therefore in this work we used four indices of instability (K, KO, Li, and TPW), whose values may very well indicate areas in an atmosphere suitable for the development of severe convection. The aim was to make their verification over greater area of Croatia. Indices were obtained by synthesis of short-term forecasts and satellite measurements, with the help of PRM method. As such, they have much better spatial and temporal resolution (if compared with radio-soundings), and cover large areas (larger than radar measurements). Specifically, in this paper, the spatial resolution of the indices was 3×3 pixels, temporal 15 minutes. Verification of the indices was made for four summer days and for one day in the winter season. Verification showed that the KO index was not efficient because of absence of 1000 hPa isobaric surface. A similar problem occurred with the generation of K index in areas where no isobaric surfaces was greater than 850 hPa. It was shown that indices intensity distribution in the morning (from 04:00 UTC to 08:00 UTC) very well coincide with the locations of convective clouds. The appearance and the area of convection were also verified with the occurrence electric discharge. Time of appearance, intensity and height at which we detected lightning showed that convection starts with its development 2 to 3 hours after the averaged indices are showing instability. Full development of convection is followed 5 to 10 hours later, thus enabling the 'forecast' of convection 2-10 hours in advance. The differences between pointed areas of instability by indices and areas of actual convection development were sometimes present, due to wind advection of atmospheric air masses with prevailing winds. The results indicated the thresholds of instability indices somewhat vary from common values. Correction for the K index would be 10 ° C, for LI index -4 ° C and +20 mm for TPW index.

convection; convective indices; lightning; severe weather; forecast

nije evidentirano