Abstract: Based on the 1959-2006 monthly and annual precipitation data from Songjiang, Donggang, Changbai, Helong, Linjiang, and Tianchi meteorological stations, and by using Morlet wavelet method, a multiple time scale analysis was made on the precipitation series from May to September (vegetation growth season), November to next April (snow season), and the annual in Changbai Mountain Region. The precipitation series was decomposed and low-frequency reconstructed at different levels by using the db5 wavelet of Daubechies wavelets, and the trend of the reconstructed series was identified and analyzed. In 1959-2006, there existed 3-6 years, 10-13 years, and 24-30 years cycles of precipitation series in vegetation growth season, 1-2 years, 5-7 years, and 17-20 years cycles in snow season, and 8-10 years, 16-20 years, and 25-30 years cycles in the annual, and the annual precipitation series showed an overall downward trend.

Key words: Changbai Mountain Region, precipitation series, wavelet analysis, wavelet variance, trend identification, winter wheat-summer maize, tillage method, leaf area, senescence enzyme activity.