Inferring faulting model using InSAR deformation differencing vector: application to the 2014 Napa earthquake

InSAR observation is widely used for mapping the coseismic deformation and estimating the earthquake source model. However, atmospheric delay noise reduces the accuracy of the mapped coseismic surface deformation, which results in the significant uncertainty of the InSAR-derived faulting model. Therefore, it is significant to mitigate the negative effect of atmospheric noise and improve the InSAR data inversion performance of earthquake source modeling. In this study, a new method of fault slip inversion is proposed, in which InSAR deformation differencing vector (IDV) is utilized as an independent constraint on source modeling instead of the original InSAR observation. According to the experimental result obtained from the simulated InSAR data, the proposed method significantly mitigates the negative effect of atmospheric noise, and the slip residual is reduced by up to similar to 65% when compared to the estimate based on the original InSAR observation. Then, the proposed method is applied to investigate the coseismic faulting of the 2014 Napa, California earthquake. It is found that the surface deformation predicted by the IDV-derived faulting model is more consistent with the GPS observation than the result obtained through the original InSAR-derived faulting model, and that the accuracy is improved by similar to 28%.

Automated summary

A new method of fault slip inversion is proposed in this study, which utilizes InSAR deformation differencing vector (IDV) as an independent constraint to mitigate the negative effect of atmospheric noise and improve the performance of earthquake source modeling. Experimental results using simulated InSAR data show that the proposed method significantly reduces slip residual by up to 65% compared to estimates based on the original InSAR observation. The method is then applied to investigate the coseismic faulting of the 2014 Napa, California earthquake, and it is found that the IDV-derived faulting model provides more accurate surface deformation prediction that is consistent with GPS observations, with an improvement in accuracy of approximately 28%.