The 1907 Karatag earthquake doublet in Central Asia: Constraints from historic seismograms and surface rupture analysis

Large historical earthquakes in intracontinental collision zones are fundamental for understanding rupture processes, fault interactions and seismic hazards. However, they remain poorly constrained due to the limited availability of instrumental data and inconsistencies in early records. Extracting reliable source parameters and reconstructing rupture dynamics of historical earthquakes therefore represent major challenges in these settings. Here, we re-examine the 1907 M_s 7.6 Karatag earthquake sequence in Central Asia using an integrated approach that combines digitized analogue seismograms, high-resolution remote sensing and tectonic geomorphology. Our analysis constrains the earthquake epicenter at ∼38.310°N, 67.007°E, with a depth of ∼17 km, a moment magnitude of M_w 7.3 ± 0.1, and strike–dip–rake angles of 190–210°, 40–60° and 80–160°, respectively. Seismograms from Japan capture two distinct surface waves separated by 21 minutes, confirming a doublet rupture. We interpret the first rupture as occurring on the NE-trending Derbent fault in the Uzbek Gissar, followed by eastward propagation onto the E-trending fault in the Tajik Gissar. The integration of waveform modeling and geomorphic evidence demonstrates how fault linkage and multi-segment rupture govern earthquake evolution in collision zones. Collectively, our results highlight the seismic hazards posed by active faults near densely populated valleys of the Gissar Mountains. Furthermore, this study shows that, when combined with modern geospatial datasets, historical seismograms can recover complex rupture processes and provide broad insights into earthquake dynamics, seismic hazard and fault kinematics in intracontinental orogens worldwide.