EQ-INSAR: A Python Package for Generating Synthetic Earthquake InSAR Deformation Data

Konrad Cieslik , Wojciech Milczarek

EQ-INSAR is an open-source Python package for generating synthetic Interferometric Synthetic Aperture Radar (InSAR) surface deformation data from earthquake sources. The package implements the Davis (1986) point source model to compute three component ground displacement from double-couple earthquake mechanisms. Then it projects the displacement onto satellite line-of-sight (LOS) geometries to produce interferometric phase maps and binary segmentation labels suitable for machine learning. EQ-INSAR supports 9 pre configured SAR satellite geometries and includes C-band, L-band, and X-band, and provides batch generation of randomized training datasets. Real data validation against the 2021 Yangbi Mw 6.1 earthquake demonstrates sub centimeter RMS fit to observed Sentinel-1 InSAR data and independent recovery of the deformation centroid to within 6km of the Global CMT solution. The software is designed for researchers developing Machine Learning based InSAR analysis tools, educators teaching InSAR interpretation, and scientists requiring synthetic benchmarks for natural and induced seismicity studies.