Exploring the Clustering Property and Network Structure of a Large-Scale Basin’s Precipitation Network: A Complex Network Approach

Abstract

Understanding of the spatial connections in rainfall is a challenging and essential groundwork for reliable modeling of catchment processes. Recent developments in network theory offer new avenues to understand of the spatial variability of rainfall. The Yellow River Basin (YRB) in China is spatially extensive, with pronounced environmental gradients driven primarily by precipitation and air temperature on broad scales. Therefore, it is an ideal region to examine the availability of network theory. The concepts of clustering coefficient, degree distribution and small-world network are employed to investigate the spatial connections and architecture of precipitation networks in the YRB. The results show that (1) the choice of methods has little effect on the precipitation networks, but correlation thresholds significantly affected vertex degree and clustering coefficient values of precipitation networks; (2) the spatial distribution of the clustering coefficient appears to be high–low–high from southeast to northwest and the vertex degree is the opposite; (3) the precipitation network has small-world properties in the appropriate threshold range. The findings of this paper could help researchers to understand the spatial rainfall connections of the YRB and, therefore, become a foundation for developing a hydrological model in further studies.