Spatiotemporal network analysis of synchronization, directional predictive dependence, and robustness in actual evapotranspiration dynamics across Iran

Long-term monitoring of actual evapotranspiration (ETa) is critical for assessing land–atmosphere water dynamics, yet optimizing monitoring networks requires a deeper understanding of spatial synchronization and predictive dependencies in water-limited environments. Utilizing GLEAM v4 satellite-derived data (1980–2023), we propose an integrated trend–topology–directionality framework to assess ETa variability across Iran. By combining long-term trend detection, complex network analysis, and Granger causality, this approach evaluates spatial connectivity to inform the design and optimization of regional environmental monitoring systems. Trend assessments (–0.03 to +0.13 mm/year) reveal rising ETa in the west/northwest and declines in the northeast, reflecting shifting hydroclimatic stress. Topological analysis of the spatial data exhibits a distinct core–periphery structure; high-degree centrality (0.001–0.43) in the central plateau (e.g., Dasht-e Kavir, Dasht-e Lut) indicates a highly synchronized core sensitive to large-scale atmospheric perturbations. Betweenness centrality metrics (≤0.03) highlight northeastern and central-eastern zones as critical spatial bridges, identifying priority locations for continuous monitoring stations. Furthermore, Granger causality isolates the northeast as a predictive source region influencing central/southeastern follower zones. Validated via spatial bootstrapping (correlation: 0.064–0.57), this framework provides a scientific basis for optimizing monitoring networks. By identifying predictive source regions and transitional nodes, the methodology provides a transferable template for environmental risk assessment and adaptive water resource management in similar large-scale, water-limited environments.