On the moisture transport regimes for extreme precipitation over North China

A novel clustering method combing the Spearman's rank spatial correlation and the spectral clustering is proposed to investigate the variability in moisture transport. Four moisture transport regimes for the extreme precipitation events (EPEs) over North China are identified through applying the new clustering method to the moisture source distributions of the 232 EPEs over North China during the rainy season (June-September) in 1979-2018. The four regimes comprise approximately 50%, 21%, 20%, and 9% of the EPEs, with counts of 115, 48, 47, and 22, respectively. Each regime exhibits distinctive moisture source pattern. In each of the four regimes, areas with higher moisture sources than the other three (meeting the statistical significance) are: the Indian Ocean and the Indochina Peninsula in regime 1, the South China Sea and the western North Pacific in regime 2, North China and its adjacent areas in regime 3, and the western North Pacific and the areas around the Maritime Continent in regime 4. Furthermore, the atmospheric circulations before the onset of EPEs manifest different patterns across the four regimes. Multiple meteorological phenomena associated with the circulation patterns are identified prior to the EPEs, e.g., the westward extension (observed in regimes 1 and 2) and eastward retreat (regime 4) of the western North Pacific Subtropical High (WNPSH), an upper-level wave train propagating from western Russia to Japan (regimes 1 and 3), extensive atmospheric river (AR) activities (regimes 1 and 4), and tropical cyclones (regime 4). The circulation patterns and the associated meteorological phenomena have critical impacts on the moisture transport and could serve as precursor signals for the EPEs.