Complex network approach for detecting tropical cyclones

Abstract Tropical cyclones (TCs) are one of the most destructive natural hazards that pose a serious threat to society, particularly those in coastal regions. In this work, we study temporal evolution regional weather conditions relation occurrence TCs using climate networks. Climate networks encode interactions among variables at different locations on Earth’s surface, and particular, time-evolving have been successfully applied phenomena comparably long time scales, such as El Niño Southern Oscillation, monsoon systems, or climatic impacts volcanic eruptions. Here, develop apply complex network approach suitable for investigation relatively short-lived TCs. We show our proposed methodology has potential identify their tracks from mean sea level pressure (MSLP) data. use ERA5 reanalysis MSLP data construct successive overlapping, short-length windows regions under consideration, where focus north Indian Ocean tropical Atlantic Ocean. compare spatial features various topological properties network, scales involved, absence presence cyclone. find measures degree clustering exhibit significant signatures striking similarities with tracks. The topology over relevant allows us obtain crucial insights into effects connectivity structure sea-level fields.