THE CHARACTERISTIC ANALYSIS OF SATELLITE IMAGES OF MESO-α SCALE STORM CLOUD CLUSTERS IN THE SOUTHEAST COAST REGION

The analysis of the convective weather phenomena (thunderstorm) occurring in the southeast coast region in China by using the GMS-5 IR, VIS and WV satellite images and usual weather charts from March to September during 1996-1999 show that the convective weather phenomena (thunderstorm) mainly take place under six different kinds of weather situations, their basic digital features of satellite images are identical, but they also can be obviously different. By analyzing the basic digital features of different satellite images in detail, we find if there are the maximum of local standard deviations and the minimum of IR minus splitting window, which is less than and equal to zero, the storm cloud clusters will reach the strongest after about four hours. So the development stages and tendency of convective weather phenomena can be understood much better, and the forecast ability of thunderstorms will be improved.