Abstract: Using global atmospheric reanalysis data ERA5 from 1981 to 2020, based on the meteorological indicators representing the degree of tea frost damage, the risk index (RI) during the spring frost monitoring period in Jiangsu province was constructed. The k-means clustering algorithm was selected to realize the spring tea frost risk zoning. Based on that, the Mann-Kendall method was applied to identify the climatic abrupt change points in each frost risk region, the impacts of climate change on the occurrence characteristics of spring tea frost in different risk areas were assessed by the daily and hourly scales, separately. The result showed that the frost risk of spring tea in Jiangsu province was characterized by "low in the southwest and high in the northeast". Light to moderate frost and extremely severe frost mainly occurred in low-risk area and high-risk area, with an average frequency of 11.91% and 15.4%, respectively. In the low-risk and high-risk areas, the average appearance times of the frost ending date showed an advance trend of 2.7d10y⁻¹ and 1.5d10y⁻¹, while RI value showed a decreasing trend of 0.024 and 0.015 per 10 years, respectively. Influenced by climate change, in terms of daily time scale, the occurrence frequency and volatility of spring tea frost at all levels during the monitoring period in low-risk and high-risk areas were significantly reduced, especially in the case of severe frost. The distribution patterns of occurrence frequency of spring tea frost at all levels changed from the type of multi peak to the type of single peak with descending trend, while still having different peak times, but mainly concentrated in the first ten days of February to the last ten days of March. From the hourly time scale, the occurrence frequency of spring tea frost at all levels basically presented a "sinusoidal" distribution pattern. After climate change, the high incidence period of mild frost was delayed for two to four hours, and that of moderate to severe frost was basically unchanged or delayed for one hour, and the high incidence period of moderate to severe frost and extra severe frost were concentrated at 5：00−7：00 and 6：00−8：00 respectively in the morning for each risk region. With the climate change, the spring tea frost in Jiangsu was characterized by "later, weaker, shorter and fewer", which still needed to be actively addressed.

Key words: Jiangsu, Spring tea, Frost, Risk assessment, Climate change