Spatial–temporal evolution characteristics and influence factors of extreme precipitation indices based on bias-corrected and gauge-measured precipitation data in the main river basins of China, 1980–2020

• Published in: Theoretical and applied climatology Vol. 155; no. 5; pp. 3563 - 3580
• Main Authors: Cheng, Qingping, Jin, Hanyu
• Format: Journal Article
• Language: English
• Published: Vienna Springer Vienna 01.05.2024; Springer Nature B.V
• Subjects: Aquatic Pollution; Atmospheric Protection/Air Quality Control/Air Pollution; Atmospheric Sciences; Bias; Circulation indexes; Climate; Climate and hydrology; Climate system; Climate variability; Climatology; Drought; Earth and Environmental Science; Earth Sciences; Evolution; Extreme weather; Floods; Gauges; Hydrologic data; Hydrology; Information flow; Personal computers; Precipitation; Precipitation data; River basins; Rivers; Slope; Slopes; Variability; Waste Water Technology; Water Management; Water Pollution Control; Wet days; China
• ISSN: 0177-798X; 1434-4483
• DOI: 10.1007/s00704-024-04826-9

Gauge-measured precipitation ( P m ) is prone to negative bias; therefore, using corrected gauge-measured precipitation ( P c ) is critical in climate and hydrology studies, although it is utilized insufficiently in existing research. This study uses P m and P c to establish extreme precipitation indices (EPIs) and compares the differences in their evolution characteristics and their responses to large-scale circulation indices (LSCIs). Theil’s slope and the modified Mann–Kendall test are used to identify the evolution characteristics of EPIs. Probabilistic condition selection and momentary conditional independence and the Liang-Kleeman information flow are adopted to evaluate the causal relationship between the EPIs and LSCIs. The main results are as follows: (1) The magnitude, intensity, and frequency of the P c -based EPIs are higher than those of the P m -based EPIs in China, except for the consecutive dry days (CDD). This result suggests that previous estimates of P m -based EPIs underestimated the magnitude, intensity, and frequency at most stations, especially in the Southwestern River where droughts and flood areas may have been underestimated. Thus, the frequency of Southwestern River droughts and floods and their sudden transitions may be higher. (2) The slope increases in the P c -based EPIs (except for the Southwestern River), and their significance levels are lower than those of the P m -based EPIs. Conversely, the slope decreases in the P c -based EPIs (except for the Southwestern River), and their significance levels (except for the Songliao River) are higher than those of the P m -based EPIs. The internal variability of the climate system and the external forcing in some basins are the reasons for the significantly increasing slopes of very wet days (R95p), simple daily intensity index (SDII), and extremely wet days (R99p). However, the significant decreasing slopes are attributed to internal climate variability. (3) Unlike the P m -based EPIs, the P c -based EPIs did not alter the causal relationship with LSCIs, indicating that P c did not change the causal relationship with LSCIs. The trend, magnitude, intensity, and frequency differ for P c and P m . Therefore, P c and P m data should be used cautiously in meteorological, hydrological, ecological, environmental, and agricultural simulations in future research.