Abstract: Increasing frequency of extreme climate events is accompanied by global warming, leading to increasing meteorological disasters. The linear propensity method and GM (1,1) catastrophic prediction method were used to analyze the meteorological data of Erguna during 1957-2017, with the aim to clarify the extreme climatic characteristics of forest-steppe ecotone in northern China. Mean annual temperature from 1957 to 2017 was -2.43 ℃, with a growth rate of 0.33 ℃·10 a^-1. The elevation of minimum temperature was the main contributor to the rising temperature. The number of frost days and freezing days, minimum temperature and its maximum value all significantly decreased, while the number of summer days, growing season length, temperature difference, maximum temperature, minimum temperature, the number of warm nights and the number of warm days increased significantly. The mean annual precipitation in Erguna in 1957-2017 was 361.6 mm, which was mainly concentrated in summer and autumn. The number of effective precipitation days decreased significantly, with a rate of 8.17 d·10 a^-1. The precipitation intensity increased significantly with a rate of 3 mm·d^-1·10 a^-1. The number of consecutive rainy days decreased significantly with a rate of 0.47 d·10 a^-1. Grey GM (1,1) model predicted that the next heavy drought event in Erguna might occur during years of 2027-2028.

Key words: sugarbeet, Na₂CO₃ stress,  , soil enzyme, soil microbe, correlation analysis, path coefficient analysis.