Combined Impacts of Climate Variability Modes on Seasonal Precipitation Extremes Over China

The joint effects of natural climate variability on the variations in seasonal precipitation extremes across China during 1961–2017 were studied based on a non-stationary GEV (generalized extreme value) model with parameters depending on multiple large-scale modes. Spatial analysis results show that the individual climate variability mode tends to have similar but weaker impacts on seasonal extremes than on mean rainfall. The combined effects of multiple large-scale modes are more likely to trigger a stronger control on the upper tail of the precipitation distribution than on mean rainfall in specific seasons. The distribution of seasonal precipitation extremes exhibits evident nonuniformity over China in different phases of the large-scale modes. Notably, the statistically significant positive responses of RX1day (maximum 1-day precipitation) and RX5day (maximum 5-day precipitation) to the El Niño–Southern Oscillation (ENSO) and Atlantic Multidecadal Oscillation (AMO) are observed at 1481 and 1416 grid points, respectively, which is more than the result (822) for SDII (simple daily intensity index). Moreover, the combined effects of ENSO and AMO on RX1day and RX5day are 10 times greater than on mean rainfall. The combined influences of three large-scale modes of climate variability on extreme precipitation events are stronger than those on mean rainfall across China in all four seasons. These phenomena suggest a closer relationship between the joint influences of multiple large-scale modes and the occurrence of seasonal precipitation extremes over China. The findings in this study will be helpful for the seasonal prediction of regional precipitation extremes and evaluating the ability of climate models to capture these teleconnection relationships.