Responses of rainfall partitioning to water conditions in Chinese forests

Abstract

The process of rainfall partitioning, which consists of throughfall, stemflow, and canopy interception, represents the initial influence of forests on the hydrological cycle. However, the impacts of water conditions on rainfall partitioning amounts remain unexplored, and the overarching patterns of this partitioning at a large scale have not been well defined. In this study, we collected data from Chinese forests to elucidate the contributing factors on rainfall partitioning amounts, and utilized machine learning techniques to predict spatiotemporal trends in rainfall partitioning amounts. Rainfall partitioning ratios were also calculated spatially to assist a better explanation for the response mechanism of the rainfall partitioning amount. The results revealed that the rainfall partitioning amounts were influenced by a combination of meteorological and vegetation factors at the site scale. Rainfall partitioning amounts were mostly influenced by rainfall among meteorological factors, and they were mainly influenced by leaf area index among vegetation factors. Through the utilization of random forest modeling, we achieved robust spatial distribution estimations, yielding R2 values of 0.90 and 0.63 for throughfall and interception quantities, respectively. Regions characterized by high levels of throughfall and canopy interception (mm) were observed in areas characterized by wet conditions (PDSI ≥ 1) in the southern part of the forests. The amounts of throughfall and canopy interception mainly showed an increasing trend, although the trend was not significant (p > 0.05). In addition, the amounts of throughfall and canopy interception were positively affected by PDSI in the majority of Chinese forests at the spatial scale. This study indicated the significance of considering water conditions when examining rainfall partitioning patterns, thereby emphasizing their critical role in the study of the hydrological cycle.