Investigating the complex multi-segment rupture of the 2025 M 6.0 Afghanistan earthquake from Sentinel-1 InSAR data
Karlee M. RiveraSubmitted August 30, 2026, SCEC Contribution #15577, 2026 SCEC Annual Meeting Poster #053
The M 6.0 2025 Afghanistan earthquake occurred in northeastern Afghanistan within the southern region of the Pamir-Hindu Kush. Afghanistan has a limited local seismic network and restricted field access, so remote sensing, particularly InSAR, is the ideal method for studying earthquakes in this area. In this study, we use Sentinel-1 InSAR data, processed with ISCE2, to model the complex rupture of the 2025 Afghanistan earthquake. The InSAR data reveal two distinct regions of surface displacement associated with this moderate-sized earthquake – a broad region with moderate uplift and a smaller, higher-amplitude region of uplift to the south. The interferometric data contained significant unwrapping errors, so we manually digitized the data by hand to produce a preliminary two-fault model to flatten the interferogram before unwrapping. To correct for the remaining phase jumps, we use the connected components mask to isolate the boundaries of discontinuous phase and correct the regions by adding or subtracting 2. We downsample the corrected data using quadtree decomposition and model fault geometry (e.g., fault length, fault width, slip) using a rectangular dislocation in an elastic half-space (Okada, 1985). We determine the best-fitting parameters through a nonlinear optimization using a Powell algorithm and multiple Monte Carlo restarts.
Our preliminary results show a near simultaneous three fault rupture involving a primary ENE-striking south dipping structure and a secondary WSW-striking north dipping structure in the south. The three faults exhibited oblique thrust faulting mechanisms, suggesting that the rupture occurred on an inward-dipping back thrust system with two primary regions of uplift: a steeper dipping northern segment at a greater depth compared to a shallower dipping thrust fault to the south. To validate the model fit to the data, we model the rupture with one, two, and three-faults. We use a statistical test, such as the Bayesian Information Criterion (BIC), to evaluate whether adding more complexity to a model is necessary. We find that the three fault model has the lowest BIC compared to the one and two-fault models, which indicates three faults are necessary to properly fit the data.
Key Words
InSAR
Citation
Rivera, K. M. (2026, 08). Investigating the complex multi-segment rupture of the 2025 M 6.0 Afghanistan earthquake from Sentinel-1 InSAR data. Poster Presentation at 2026 SCEC Annual Meeting.
Related Projects & Working Groups
Tectonic Geodesy
