Abstract: Mesoscale convective systems (MCSs) can cause heavy rain and other severe weather events at summer. Satellite IR images with high spatial and temporal resolution is one of main approach to monitor behaviors of the MCSs. In present paper, the smallest scale of meso β scale was defined as 1.0 to 3.0 latitudes according to actual situation in Xinjiang. The hourly GMS-5 meteorological satellite infrared images data from 1998 to 2002 in Xinjiang were processed by infrared satellite image enhancement and image overlay processing technology in order to analyze the behaviors of MCSs. The results showed as follows: (1) in this five years, the MCS in meso β scale did not appear, the MCS in meso β scale was happed 111 times. MCS has uneven characters: frequently occurs in 1998, 2000 while rare occurs in 2002. MCSs are concentrated in two major areas: one is the windward slop of Tianshan Mt. in the eastern of Ili River Vally, another Keping at the lee side of West Tianshan Mt. There occur less frequently in Gurbantunggut Desert while more in Taklimakan Desert; (2) Most of MCSs occur at end of spring and the summer while it rarely occur at autumn and winter. In June, the number of occurrences of MCSs accounts for 34 % of the total while in May and July respectively 25 % and 15 %, this fact is related to the high incidence season of the severe weather such as duststorm and hail in this area; (3) Most of MCSs possesses nocturnal feature, and it occurs from noon to midnight, and dissipates before dawn, the duration is 4 - 6 h; (4) The shape of MCSs in Xinjiang consist of round and ovalshaped, and each of them is accounts respectively for 50 % of the total. Moreover, the MCS is developed from the convection bubble, so scale of the majority of MCS is small; (5) the MCS example in different area reveals the diversity of blackbody temperature distribution in cloud top and MCS evolution process.

Key words: infrared image, mesoscale convective system, geographic distribution, life cycle