Application of an improved multifractal detrended fluctuation analysis approach for estimation of the complexity of daily precipitation

Affected by climate change and high-intensity human activities, the precipitation series is characterized by nonlinear complex fluctuations. In order to accurately quantify the precipitation complexity, the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) method was introduced to address the shortcomings of the traditional multifractal detrended fluctuation analysis (MFDFA), and the CEEMDAN-MFFA method was proposed to explore the spatiotemporal variability and possible causes of precipitation complexity in Northeast China (NEC). The results showed that the traditional MFDFA method generally overestimated precipitation complexity. However, the CEEMDAN-MFFA method improved the accuracy and reliability of precipitation complexity analysis. Daily precipitation series in NEC were characterized by multifractal with high uncertainty and anti-persistence. The lowest precipitation complexity was 0.573, which appeared in Qian'an, Jilin Province, and the highest precipitation complexity was 1.108, which appeared in the Boketu, Inner Mongolia. The cities with high complexity of daily precipitation are sporadically distributed in the three provinces of Heilongjiang, Jilin and Liaoning. The average complexity is similar to 0.892, accounting for 22.5% of the total administrative districts. The complexity of daily precipitation showed a statistically insignificant downward trend with time. The precipitation dynamics structure in the southeast and northwest of the study area was complex, and the precipitation was less predictable than other regions. However, the precipitation dynamics structure was relatively simple and the precipitation was easily predictable in the northeastern, central and southwestern regions. Elevation, topographic relief, changes in water area, forest area and agricultural production were the main influencing factors for the differences in precipitation complexity. Results also revealed that increasing the area of forestland by means of protecting and breeding forests, via regulation of the forest ecosystems to alleviate the complex precipitation process will become a reliable approach to cope with nonstationary precipitation.

Automated summary

The study revealed that in Northeast China, daily precipitation exhibits high uncertainty and antipersistence multifractality. The CEEMDAN-MFFA method proved to be more accurate and reliable in analyzing precipitation complexity in the region.