Network-based exploration of basin precipitation based on satellite and observed data

Adequate and efficient precipitation data is a major concern due to its spatiotemporal variability and topographic and climatic factors. Satellite-based products are an alternative for a reliable precipitation estimate in basins having a complicated topography and diverse climate zones. Satellite products with global coverage and continuous data are freely available; however, understanding spatial connections is essential for reliable hydrological applications. In this study, complex network concepts like clustering coefficient, degree, degree distribution, average neighbour and architecture employed to investigate spatial connections in a basin. We also identified influential grid points in the precipitation network using weighted degree betweenness. Our results reveal that the correlation method does not significantly affect the network topology. However, the correlation threshold influences the spatial distribution of the clustering coefficient and degree values of precipitation network. The spatial distribution of clustering coefficient and degree indicated an inverse relationship independent of similarity measures and correlation thresholds. The architecture of precipitation based on satellite and observed data shows small-world behaviour for the certain correlation threshold range. Our findings unravel spatial precipitation connections and provide a way for hydrological applications in further research.