Linking dynamic earthquake rupture to tsunami modeling for the Húsavík-Flatey transform fault system in North Iceland

Earthquake generated tsunamis are generally associated with large submarine events on dip-slip faults, in particular on subduction zone megathrusts (Bilek and Lay, 2018). Submerged ruptures across strike-slip fault systems mostly produce minor vertical offset and hence no significant disturbance of the water column. For the 2018 Mw 7.5 Sulawesi earthquake in Indonesia, linked dynamic earthquake rupture and tsunami modeling implies that coseismic, mixed strike-slip and normal faulting induced seafloor displacements were a critical component generating an unexpected and devastating local tsunami in Palu Bay (Ulrich et al., 2019), with important implications for tsunami hazard assessment of submarine strike-slip fault systems in transtensional tectonic settings worldwide. 

We reassess the tsunami potential of the ~100 km Húsavík Flatey Fault (HFF) in North Iceland using physics-based, linked earthquake-tsunami modelling. The HFF consists of multiple fault segments that localise both strike-slip and normal movements, agreeing with a transtensional deformation pattern (Garcia and Dhont, 2005). The HFF hosted several historical earthquakes with M>6. It crosses from off-shore to on-shore in immediate proximity to the town of Húsavík. We analyse simple and complex fault geometries and varying hypocenter locations accounting for newly inferred fault geometries (Einarsson et al., 2019), 3-D subsurface structure (Abril et al., 2020), bathymetry and topography of the area, primary stress orientations and the stress shape ratio constrained by the inversion of earthquake focal mechanisms (Ziegler et al., 2016).

Dynamic rupture models are simulated with SeisSol (https://github.com/SeisSol/SeisSol), a scientific open-source software for 3D dynamic earthquake rupture simulation (www.seissol.org, Pelties et al., 2014). SeisSol, a flagship code of the ChEESE project (https://cheese-coe.eu), enables us to explore simple and complex fault and subsurface geometries by using unstructured tetrahedral meshes. The dynamically adaptive, parallel software sam(oa)²-flash (https://gitlab.lrz.de/samoa/samoa) is used for tsunami propagation and inundation simulations and solves the hydrostatic shallow water equations (Meister, 2016). We consider the contribution of the horizontal ground deformation of realistic bathymetry to the vertical displacement following Tanioka and Satake, 1996. The tsunami simulations use time-dependent seafloor displacements to initialise bathymetry perturbations. 

We show that up to 2 m of vertical coseismic offset can be generated during dynamic earthquake rupture scenarios across the HFF, which resemble historic magnitudes and are controlled by spontaneous fault interaction in terms of dynamic and static stress transfer and rupture jumping across the complex fault network. Our models reveal rake deviations from pure right-lateral strike-slip motion, indicating the presence of dip-slip components, in combination with large shallow fault slip (~8 m for a hypocenter in the East), which can cause a sizable tsunami affecting North Iceland. Sea surface height (ssh), which is defined as the deviation from the mean sea level, and inundation synthetics give an estimate about the impact of the tsunami along the coastline. We further investigate a physically plausible worst-case scenario of a tsunamigenic HFF event, accounting for tsunami sourcing mechanisms similar to the one causing the Sulawesi Tsunami in 2018.