根据“倒春寒”气象指标国家标准（GB/T 34816-2017），利用金衢盆地主要城市历史气象资料，整理1981—2020年“倒春寒”个例并计算得到每次“倒春寒”过程的气象指标K值。应用统计分析方法和小波分析方法，分析了金衢盆地“倒春寒”频次和强度变化特征及大气环流背景。结果表明，近40 a金衢盆地“倒春寒”发生频繁，但以轻度“倒春寒”为主，重度“倒春寒”仅占发生次数的3.4%;“倒春寒”气象指标K值在3上下波动;衢州市“倒春寒”次数和气象指标K值分别存在明显的3 a、7 a、19 a和4 a、8 a、20 a左右的年际和年代际周期变化规律;“倒春寒”较强年受贝加尔湖地区强冷空气东移南下影响，槽后冷平流不断渗透影响江南地区，导致金衢盆地气温异常下降，发生较严重“倒春寒”气象灾害。

According to the national standard（GB/T 34816-2017） for the meteorological indicators of late spring cold, using the historical temperature data of major cities in the Jinqu basin, sort out the cases of late spring cold from 1981 to 2020 and calculate the K value of the meteorological index for each late spring cold. The statistical analysis method and wavelet analysis method were used to analyze the frequency and intensity change characteristics of late spring cold in the Jinqu basin and the atmospheric circulation background. The results showed that the late spring cold in the Jinqu basin occurred frequently in the past 40 years, but the mild late spring cold accounted for a large proportion, and the severe late spring cold only accounted for 3.4%. The index K of late spring cold showed a significant wave dynamic trend, and the K value fluctuated around 3. There were obvious interannual and interdecadal cycle changes of late spring cold number and meteorological index K of 3 a, 7 a, 19 a and 4 a, 8 a, and 20 a in Quzhou. The strong late spring cold was affected by the strong cold air moving eastward and southwards in the Baikal region. The cold advection after the trough continues to infiltrate the south of the Jiangnan region, resulting in an abnormal temperature drop in the Jinqu basin and a severe late spring cold weather disaster.