Unevenness of temporal distribution of precipitation over the Yangtze River basin and its main influencing factors

Understanding the unevenness of precipitation distribution within a year is important for water resources management and agricultural production. In this study, spatiotemporal variations of precipitation unevenness in the Yangtze River basin (YRB) during 1961-2020 and its main influencing factors are investigated by the relative entropy (RE) and rotated empirical orthogonal function. The timings of different percentiles of annual total precipitation show that the temporal distribution of precipitation in the YRB is uneven. The unevenness of precipitation distribution exhibits a large spatial variability with a high-value centre in the northeast and a low value centre in the northwest of the YRB, and an upward trend in almost the entire YRB. Precipitation becomes more uneven with the decrease of annual number of wet days and low-intensity precipitation amount and the increase of high-intensity precipitation amount during the study period. Precipitation unevenness of different regions is modulated by the different combinations of different seasonal climate indices with specific time lags. For example, in the south-central YRB, precipitation unevenness generally tends to be enhanced by the positive Tropical Northern Atlantic Index in current summer and sea surface temperature anomalies over South China Sea (SCS) in previous summer, and negative North Atlantic Oscillation in current winter, while in the northeast YRB, be weakened by negative Quasi-Biennial Oscillation in current spring, Atlantic Meridional Mode in current summer and positive SCS in current autumn. Besides, the identified significant climate indices can partly explain the temporal changes of precipitation unevenness, and their explanatory ability is more powerful in the south-central, southeast and northwest YRB than in the north-central and northeast YRB, which reveals a strong negative correlation with the degree of precipitation unevenness.