A catastrophe identification method for rainfall time series coupled sequential Mann-Kendall algorithm and Bernaola Galvan algorithm: a case study of the Qinglong River watershed, China

Abstract

The identification of rainfall catastrophe characteristics is important for rainfall consistency testing in hydrostatistical analysis. In this study, a new classification framework method (trend, mean and change-rate catastrophe) coupled Sequential Mann-Kendall (SQ-MK) algorithm and Bernaola-Galvan heuristic segmentation (B-G) algorithm was proposed, and a rainfall catastrophe identification scheme was carried out for the Qinglong River watershed, Northern China. Meanwhile, the accuracy of catastrophe identification was improved by the robustness of the algorithm and the parameter optimization method of the B-G algorithm. Results revealed that (1) the most significant point of trend catastrophe in the Qinglong River watershed was in 1997. The trend catastrophe identification based on the SQ-MK algorithm was sensitive to the length of time series. (2) The sensitivity of parameter P0 (Range value (R) = 2.889) in the B-G algorithm was greater than that of parameter l0 (R = 0.333). The mean catastrophe points for the Qinglong River watershed were in 1997, 2002 and 2009. (3) The mean catastrophe identification based on the B-G algorithm was insensitive to the length of time series. (4) There was no change-rate catastrophe point in the Qinglong River watershed. Trend catastrophe and mean catastrophe did not necessarily lead to change-rate catastrophe.