Jump to Main Content
Identification and application of the most suitable entropy model for precipitation complexity measurement
- Zhang, Liangliang, Li, Heng, Liu, Dong, Fu, Qiang, Li, Mo, Faiz, Muhammad Abrar, Khan, Muhammad Imran, Li, Tianxiao
- Atmospheric research 2019 v.221 pp. 88-97
- altitude, atmospheric precipitation, drought, entropy, floods, grasslands, models, time series analysis, uncertainty, water management, wavelet, China
- Precipitation complexity measurement is often overlooked in precipitation time series research. Entropy, as a measure of system complexity, can be used to diagnose the complexity of precipitation. However, it is difficult to judge the applicability of different theoretical entropy models for solving precipitation uncertainty problems. This paper introduces the distinction degree theory and the serial number sum theory to screen the optimal entropy model for precipitation complexity measurement. The optimal entropy model was used to analyze the spatiotemporal differences of monthly precipitation complexity in Heilongjiang Province, China. Possible influencing factors of precipitation complexity were also examined. The results indicated that in the complexity measurement of precipitation based on entropy theory, the stability and reliability of sample entropy was higher than those of fuzzy entropy, wavelet entropy and permutation entropy. The complexity of monthly precipitation in the selected study area significantly increased with time. The average complexity of monthly maximum precipitation, monthly average precipitation and monthly minimum precipitation were 0.665, 0.622 and 0.545, respectively, and their tendency change rates were 0.070/decade, 0.055/decade and 0.038/decade, respectively. The areas with high monthly precipitation complexity were concentrated in the central, eastern and northwest parts of the study area, and the precipitation was less predictable. Monthly precipitation in the southwest was less complex and more predictable. The highest monthly precipitation complexity was 1.012, at Hulin station, and the lowest was 0.510, at Mingshui station. The increasing complexity of monthly precipitation in the province was strongly related to local industrial and agricultural production. The superposition effects of altitude, topographic relief, change in grassland area and agricultural production formed the basic pattern of spatial differences in monthly precipitation complexity. The results may provide a scientific guidance for regional precipitation predictability measurement, effective assessment of droughts and floods, and water resources management.