This study aims to identify earthquake clustering patterns in the Northern Lombok–Flores Back-Arc Zone using the Agglomerative Clustering method. Seismic data were obtained from the United States Geological Survey (USGS) covering the period 1970–2021, using latitude, longitude, depth, and magnitude as clustering variables. A sliding window approach was applied to capture the temporal evolution of seismic activity. The results indicate the formation of three stable clusters that consistently represent different earthquake depth levels across all analyzed time windows. Although fluctuations in the occurrence of large-magnitude earthquakes were observed in several windows, these variations did not significantly affect the clustering structure. This study demonstrates that the combination of Agglomerative Clustering and a sliding window approach is effective for identifying depth-based earthquake clustering patterns and describing their temporal dynamics
Copyrights © 2025