Abstract: Based on the data of annual maximum precipitation of different durations from 1958 to 2022 at Kunming Climate Reference Station, combined with short-duration precipitation index analysis, trend analysis, wavelet analysis and extreme value analysis, the basic evolution characteristics of annual maximum precipitation of different durations in Kunming were revealed, the extreme value distribution of precipitation of different durations was fitted and the result of recurrence period was calculated. The results indicate that: (1) The values of all precipitation indices for different durations of annual maximum precipitation in the Kunming area are lower than those in regions such as Zhengzhou, Beijing, and Zhejiang. This is attributed to the differences in climatic endowments across various regions. (2) The annual maximum precipitation of different durations in the Kunming area exhibits significant interannual variability, with an overall weakly increasing trend. Except for the two time periods of 120 minutes and 180 minutes, where the annual maximum precipitation shows significant periodic changes from 11 to 14 years, the periodic changes in precipitation in other time periods are not very significant. (3) Statistical inspection indicators and key data point tests show that the Gumbel distribution performs favorably in Kunming and can be used for calculating the precipitation intensity for any recurrence period of different durations in Kunming. The research findings can provide a basis and reference for understanding, predicting, and warning against heavy precipitation in the Kunming area, meteorological services, and urban disaster prevention and mitigation. They also offer insights for similar work in other regions.

Key words: Kunming, different duration, precipitation extremes, variation characteristic, recurrence period