To date, most observational earthquake research relies on ground motions recorded by seismometers. These instruments are typically installed in proximity to active faults, as the most valuable observations are those obtained very close to earthquake epicenters: they provide the most coherent view of source processes and allow for early detection of large earthquakes and monitoring of small ones. However, there is a severe observational gap: the vast majority of seismometers are located on-land, while the largest earthquakes, and most tsunami generating earthquakes, occur underwater. Existing technologies to overcome this observational gap, e.g., ocean-bottom seismometers (OBS), are very costly and thus not widely implemented. The lacking ocean-bottom monitoring hinders the ability to conduct underwater seismological research. This is especially critical for hazard mitigation tasks such as providing earthquake early warning (EEW) (e.g.,

Earthquake detection capabilities using DAS are similar to those of broadband instruments.• Detection capabilities are mainly a function of the recorded noise, cable response and apparent velocity.

The energy budget during the propagation of tensile fractures typically includes two major energy dissipation modes: the fracture energy to create a new interface and the radiated energy. Since seismic hazard is related to the amount of seismic energy released, it is important to evaluate precisely the energy budget during fracture propagation and see in particular if it is constant or not. However, two aspects are typically limiting a precise estimate of the energy budget: first, the measurement of the nonseismic energy release and second, the rock heterogeneity-like asperities or barriers. We conducted here laboratory experiments, using an analog material (Polymethyl methacrylate (PMMA)), of a stable mode I interfacial crack propagation close to brittle-creep transition through a heterogeneous interface for different macroscopic rupture velocities to evaluate carefully their energy budget. Both acoustic and optical advances of the crack front were measured simultaneously, providing precise estimates of both types of dissipated energy. We computed the radiation efficiency R (ratio of the radiated energy to the available energy for driving the fracture) and observed a nonlinear increase of R with the average fracture propagation velocity v over 2 orders of magnitude independently of the initial quenched disorder. The experimental observations are supported by a model based on the fluctuations of the local rupture velocity induced by the crack front pinning on local asperities which leads to R ∝ v 0.55 . We discuss implications for slow shear rupture modes, seismicity rate evolution, and induced seismicity.

We monitor optically the propagation of a slow interfacial mode III crack along a heterogeneous weak interface and compare it to mode I loading. Pinning and depinning of the front on local toughness asperities within the process zone are the main mechanisms for fracture roughening. Geometrical properties of the fracture fronts are derived in the framework of self-affine scale invariance and Family-Vicsek scaling. We characterize the small and large scale roughness exponents ζ_{-}=0.6 and ζ_{+}=0.35, the growth exponent at large scale β_{+}=0.58, and the power-law exponent of the local velocity distribution of the fracture fronts, η=2.55. All these analyzed properties are similar to those previously observed for mode I interfacial fractures. We also observe a common power-law decay of the probability distribution function of avalanche area. We finally observe that amplitude of front fluctuations, local rupture velocity correlation in time, and larger size of events highlight more dynamically unstable behavior of mode III crack ruptures.