Abstract:The unconditional and conditional probability density functions of daily precipitation are estimated by use of the 1961-2000 summer daily precipitation data of 174 stations in China,and the probability distributions of maximum daily precipitations,and of the maximum daily precipitation greater than or equal to 10 mm,25mm,and 50mm within 10,and 20 days for each station are derived,respectively.The main results show that the differences of the Gamma distribution shape parameter α of daily precipitation under no condition,and the condition of a wet/dry preceding day at different stations are not too large,and all α are smaller than 1.While the scale parameter β of the Γ distribution,to the overwhelming majority of stations,the unconditional β is greater than the dry preceding day β0 and less than the wet preceding day β1.Both unconditional and conditional μ and σ2 in the southeast coastal region are greater than those in the northwest inland area;and for the overwhelming majority of stations,the unconditional μ are greater than the dry preceding day μ0 and less than the wet preceding day μ1,and the same to the variance of Γ distribution.Γ distribution can accurately approximate to the sample frequency distributions of daily precipitation under the conditions of a wet or dry preceding day.The probability distributions of the daily maximum precipitation greater than 10mm,25 mm,and 50 mm within 10 days,and 20 days in the southeast coastal region are all greater than those in the northwest inland area,respectively,approaching to the geographic distributions of the sample frequency of the same rainfall order.The high value area of probability lies in the west of Guangxi and the east of Yunnan,and the secondary high value area in the south of Anhui and the north of Jiangxi.