Complex network analysis of extreme precipitations in North America

Abstract

Improving climate models for predicting water-related hazards such as floods and managing water resources depends on identifying the complex patterns of extreme precipitation events (EPEs). In this regard, we studied EPEs in North America using complex network analysis. We generated the EPEs network from the gauge-based daily precipitation dataset from the Climate Prediction Center database. The geographical grid points serve as the network's nodes, while the associations among them, captured using a non-linear method called event synchronization, represent the network's links. We determined the spatiotemporal patterns of the dependencies of EPEs through network measures such as degree centrality, mean geographical distance, betweenness centrality, clustering coefficient, and long-ranged directedness. We found hubs—locations important in propagating EPEs and teleconnections—in areas such as Montana, Wyoming, Alberta, and Saskatchewan in the summer and the West Coast and eastern North America in the winter. Our work demonstrated that EPEs are more spatially coherent during winter than summer across the continent. We identified certain locations in Utah, Colorado, South Dakota, northern Mexico, southeast British Columbia, and northern Quebec that play a crucial role in the long spatial propagation of EPEs from one region to another throughout the summer, while regions in the West Coast, southern Colorado, New Mexico, Wisconsin, central Alberta, etc., are dominant in the winter. Additionally, we uncovered atmospheric moisture pathways for EPEs in both seasons on the continent. Finally, we showed that EPEs networks are sensitive to 𝜏[subscript 𝑚𝑎𝑥], a parameter of the ES method. As such, it should be defined relative to the atmospheric phenomena of interest. The insights from this study revealed synchronization patterns of EPEs across North America and advanced our understanding of complex relationships and interdependencies between topography and precipitation patterns. Further, analytics provided here can serve as a basis for improving the forecasting of hydrologic extremes and associated risks.