Exploring latent heat flux anomalies for seismic activity detection: insights from RST methodology and case studies
摘要
As a sudden and catastrophic natural disaster, earthquakes pose severe threats to human lives and property. In recent years, pre-seismic thermal anomalies, particularly latent heat flux (LHF) anomalies, have gained increasing research interest due to their potential correlation with the earthquake preparation process. However, existing studies are still constrained by insufficient spatiotemporal resolution and poor data consistency. This study systematically analyzes pre-seismic anomalies associated with four major earthquakes in China—the Wenchuan, Lushan, Ludian, and Yushu events—by integrating multi-resolution LHF products (GLEAM, PML, ETMonitor) with the Robust Satellite Technique (RST). Results indicate that significant LHF anomalies emerge approximately 1–2 months prior to the Wenchuan, Lushan, and Ludian earthquakes, clustering within 20 km of epicenters and fault zones, with abnormal pixel proportions reaching 17–50% before the mainshocks and declining to below 14% afterward. Sensitivity analysis reveals that an RST threshold of 1.5 achieves optimal balance among hit rate and false alarm rate. Land use analysis shows anomalies primarily occur over grasslands, shrublands, and croplands—accounting for the majority of abnormal pixels—while higher-resolution datasets (ETMonitor, PML) detect supplementary anomalies in forested areas and water bodies, highlighting the role of spatial resolution. In contrast, the Yushu earthquake exhibits a weaker signal, with abnormal pixel proportions limited to 3–15%, likely due to cold, arid environmental constraints. Overall, this study validates the effectiveness of combining LHF data with the RST method for detecting potential seismic precursors and provides new observational evidence and practical insights for remote sensing–based earthquake monitoring.