Abstract: Based on rice biological characteristics, circadian temperature rhythm and temperature coefficient, taking accumulated equivalent temperature, negative accumulated equivalent temperature and heat matching index as index, the identification method for cold damage year was studied. By using moving average method on 21 years, departure percent was obtained for three factors in Heilongjiang province from 1971 to 2016, and the identifiable ability was analyzed. The results showed that identification ability was stronger, when the departure percent of every factor decreased more. This identification model for cold damage based on integrative departure percent was established by using weighted membership model. When departure percent was above ?5%, named as zero cold damage year, when departure percent was above ?10% and below ?5%, named as slight cold damage year, when departure percent was above ?15% and below ?10%, named as moderate cold damage year, and when departure percent was below ?15%, named as severe cold damage year. Integrative departure percent could indicate cold damage during rice growing season, and could reflect different degrees of cold damage, and totally indicated typical cold damage years. Cold damage performed a feature of outbreak in Heilongjiang province, for example in 1971, 1972, 1976, 1981, 1983, 1992 and 2009. In these years, cold damage extended over extensive regions, and the damage degree was more severe, the other damage year was slight degree or regional damage, including 1986, 1987, 1989, 1993, 1995 and 2003.

Key words: Delayed-growth type cold damage, Accumulated equivalent temperature, Departure percent, Cold damage year, Cold damage degree