Research on the Degree of Non-Stationarity in Extreme Precipitation in the Continental United States

The non-stationarity of identifying hydrological extreme sequences is essential for understanding the patterns of hydrological systems and conducting reasonable risk assessments. Currently, research on detecting non-stationarity primarily determines whether the parameters of the sequence change through statistical tests. The study introduces the concept of distribution change to detect non-stationarity and proposes the Cumulative Distribution function Change Indicator (CDCI) to quantify the degree of non-stationarity in extreme sequences. Based on Hourly Precipitation Data from 102 U.S. Weather Stations, 1975–2021. Comparison with other methods commonly used for non-stationarity identification validates the reasonableness of CDCI. Additionally, it explores the relationship between non-stationarity, distribution, and return period based on CDCI. The research results demonstrate that: (1) Distribution change can reflect the non-stationarity of sequences, and CDCI is more sensitive compared to conventional non-stationarity identification methods; (2) By combining CDCI with return period change, the response of distribution to non-stationarity can be identified, including the location and degree of distributional change. The study demonstrates the feasibility and validity of distributional change being used as a measure of non-stationarity in hydrological extreme sequences. Furthermore, the results reveal the possible relationship between non-stationarity, distribution change, and return period change.