Spatiotemporal connections in high precipitation events in Iran: Application of complex networks

Mahnoor Roohinia, Banafsheh Zahraie

Study region
A relatively large area covering east and west Asia, north Africa, and Europe.
Study focus
This study examines the complex correlation patterns of high precipitation events in Iran and the rest of the study region. For this purpose, the Complex Networks Theory is used to find the links between high precipitation events in Iran and the rest of the study area with different time lags. The work pinpoints persistent teleconnection routes as well as changes in these routes, possibly due to global warming. Monthly GPCP precipitation values from 1980 to 2019 are used in the analysis. Precipitation events exceeding the 70th percentile in each grid cell are identified and analyzed across twenty-one annual sliding windows. The sliding windows detect the effects of global climate change on the teleconnections of high precipitation events with different lag times. Event Synchronization (ES) and Event Coincidence Analysis (EC) are utilized as complementary methods for creating adjacency matrices. Network structures are also delineated through clustering coefficient as well as betweenness centrality measures.
New hydrological insights for the region
The results of this study demonstrate that with a zero-time lag, large-scale local precipitation clusters develop synchronously over both eastern and western Iran. In contrast, when including a 2–3-month delay, precipitation clustering becomes dominant in southeast Iran. Moreover, the betweenness centrality analysis indicates a shift in critical moisture transition nodes. These nodes move from synchronous regions in southeast Iran and its border with Pakistan toward delayed routes passing through northwest Iran, southern India, and the Yemen–Saudi Arabian border. The combination of ES and lagged EC offers a dual approach for identifying both synchronous and lagged precipitation teleconnections, thereby providing a robust opportunity for the seasonal forecasting of high-precipitation events. Identifying zones with high clustering and centrality across multiple temporal scales yields new insights for developing effective early flood warning systems and effective mid-term water resources planning and management.