利用1960年~2017年贵州梵净山区域10个国家观测站的逐日降水资料，采用EOF分解法、线性回归、趋势检验、距平累积、相关分析方法，分析该区域年雨日、不同等级雨日的空间变化特征和气候变化趋势，以及不同等级雨日与降雨量的关系。结果得出：总雨日的空间典型场特征中76.1%为全区一致型；近58a中，总雨日呈中部少雨日带、东西两边多雨日带分布，全区域均为一致性减少趋势，多雨日区减小趋势小，少雨日区减小趋势明显，总雨日的年代际分布经历了偏多–偏少–偏多–偏少的过程；年小雨日数呈东多西少分布，年中雨日数呈西北部向东南部增多分布，年大雨日数呈西南少东部多分布，年暴雨日数呈西南部少、东北 + 西北多的人字形分布。年雨日数与年降雨量近58a的曲线变化相似，在1982年和2002年有明显偏多向偏少的转折，20世纪之后雨日数下降的趋势较降水量更为明显；各级雨日数均与年降水量呈正相关关系，大雨日数与雨量的线性相关系数最大，小雨日数与雨量的线性相关系数最小。
Based on the daily precipitation data of 10 national observatories in Fanjingshan region of Guizhou from 1960 to 2017, the spatial variation characteristics and climate change trends of annual rainy days and different grades of rainy days in this region were analyzed by EOF decomposition method, linear regression, trend test, anomaly accumulation and correlation analysis method, and the relationship between different grades of rainy days and rainfall was also analyzed. The results show that 76.1% of the spatial typical field characteristics of total rainy days are consistent in the whole region; in the past 58 years, the total rainy days are distributed in the middle and the east and west rainy days, with a consistent decreasing trend in the whole region, with a small decreasing trend in the rainy days and an obvious decreasing trend in the less rainy days. The interdecadal distribution of the total rainy days has experienced more - less - more - less; the annual light rain days are less in the east and west, while the annual rain days increase from northwest to southeast. The annual heavy rain days are less in the southwest and more in the east, and the annual heavy rain days are less in the southwest and more in the northeast + northwest. The curve of annual rain days is similar to that of annual rainfall in recent 58 years. In 1982 and 2002, there was an obvious turning point from more to less. After the 20th century, the trend of rain days decreased more obviously than that of precipitation. There is a positive correlation between rain days at all levels and annual precipitation. The linear correlation coefficient between heavy rain days and rainfall is the largest, while the linear correlation coefficient between light rain days and rainfall is the smallest.
