Seismo-geodetic modelling of complex faulting earthquake

Reliable seismic hazard assessments, active tectonics and earthquake physics studies depend on accurate and robust earthquake source models. A simplifying assumption in routinely estimated earthquake source models is that the event can be described by a single point source. However, not only large megathrust earthquakes often rupture on different fault segments, but even smaller magnitude events can show complex faulting mechanisms. In these cases, a single fault model oversimplifies the source process. Different data types from local seismic networks, regional waveforms, stations at teleseismic distances and geodetic data such as space-born radar imagery (e.g. InSAR) or GNSS provide complementary information about the earthquake's source. Thus, simultaneous inversions for multiple fault source mechanisms using these data sets are highly beneficial for accurate descriptions of source processes. Moreover, while source studies based on seismic data often neglect the 3-D Earth structure, recent research highlighted the importance of such effects. We present a new joint seismo-geodesy inversion method for the simultaneous determination of multiple fault source solutions. Our technique takes 3-D Earth structure effects fully into account when modelling seismic data and uses a Monte Carlo method to explore the model space. A series of synthetic tests is performed to validate the use of the algorithm in order to investigate complex faulting events with multiple sub-faults. We then study the 21st February 2008, Mw 6.0 Wells earthquake in Nevada, USA using local seismic, teleseismic and InSAR data to obtain its source parameters and associated uncertainties.The synthetic tests show that the algorithm performs well and highlight its advantages at correctly retrieving multiple fault source parameters compared to an iterative inversion approach. A two-fault solution for the Wells event leads to an improvement in the data fit compared to a single fault source inversion and seems to match the geometry of the aftershocks. The estimated source parameters are highly beneficial to explore earthquake physics, notably to constrain the earthquake's stress drop and energy budget.