Earthquake damage patterns resolve complex rupture processes

Large continental earthquakes activate multiple faults in a complex fault system, dynamically inducing co-seismic damage around them. The 2016 Mw 7.8 Kaikoura earthquake in the northern South Island of New Zealand has been reported as one of the most complex continental earthquakes ever documented1, which resulted in a distinctive on and off-fault deformation pattern. Previous geophysical studies confirm that the rupture globally propagated northward from epicenter. However, the exact rupture- propagation path is still not well understood because of the geometrical complexity, partly at sea, and the possibility of a blind thrust. Here we use a combination of state-of- the-art observation of surface deformation, provided by optical image correlation, and first principle physics-based numerical modeling to determine the most likely rupture path. We quantify in detail the observed horizontal co-seismic deformation and identify specific off-fault damage zones in the area of the triple junction between the Jordan, the Kekerengu and the Papatea fault segments. We also model dynamic rupture propagation, including the activation of off-fault damage, for two alternative rupture scenarios through the fault triple junction. Comparing our observations with the results from the above two modeled scenarios we show that only one of the scenarios best explains both the on and off-fault deformation fields. Our results provide a unique insight into the rupture pathway, by observing, and modeling, both on and off-fault deformation. We propose this combined approach here to narrow down the possible rupture scenarios for large continental earthquakes accompanied by co-seismic off-fault damage.

💡 Research Summary

The paper addresses the long‑standing problem of determining the exact rupture pathway of the 2016 Mw 7.8 Kaikoura earthquake, which is renowned for its extraordinary complexity involving multiple intersecting faults. While previous seismological studies have established that the rupture propagated northward from the epicenter, the precise geometry—especially through the offshore segment and the triple junction formed by the Jordan, Kekerengu, and Papatea fault strands—remained ambiguous. To resolve this, the authors combine two state‑of‑the‑art techniques: high‑resolution surface deformation mapping using optical image correlation, and physics‑based dynamic rupture simulations that explicitly incorporate off‑fault damage.

First, pre‑ and post‑event aerial and satellite imagery were co‑registered and processed with a correlation algorithm capable of detecting horizontal displacements on the order of a few centimeters. The resulting displacement field reveals not only the expected on‑fault slip but also a series of distinct off‑fault damage zones (ODZs) clustered around the triple junction. These ODZs are characterized by anomalous strain orientations and higher strain magnitudes than the surrounding crust, indicating localized brittle failure away from the main fault planes.

Second, the authors construct a three‑dimensional, non‑linear dynamic rupture model that faithfully reproduces the known fault geometry, including curvature, step‑overs, and a hypothesized blind thrust offshore. The model uses first‑principles constitutive laws (elastic‑plastic rheology, rate‑and‑state friction, and velocity‑weakening) to govern rupture propagation and to trigger off‑fault damage when stress exceeds a prescribed failure envelope. Two competing rupture scenarios are simulated: (A) a sequential propagation from the Jordan fault to Kekerengu and then to Papatea, and (B) a direct jump from Jordan to Papatea followed by later activation of Kekerengu.

Both simulations generate synthetic on‑fault slip distributions and patterns of off‑fault damage. When compared with the observed displacement field, scenario A reproduces the magnitude and direction of the on‑fault slip as well as the spatial arrangement of the ODZs. In contrast, scenario B fails to match the observed ODZ locations and predicts slip orientations that are inconsistent with the optical data. The authors therefore conclude that the rupture most likely followed the sequential path (Jordan → Kekerengu → Papatea), with a rapid acceleration through the triple junction that induced intense off‑fault fracturing.

The study provides several key insights. It demonstrates that off‑fault damage, often overlooked in traditional seismological analyses, can serve as a powerful tracer of rupture dynamics in complex fault networks. The integration of high‑resolution optical correlation with physics‑based modeling offers a quantitative framework for discriminating between competing rupture hypotheses. Moreover, the work highlights the importance of incorporating realistic fault geometry—including hidden offshore structures—into dynamic rupture simulations to capture the full spectrum of co‑seismic deformation. The authors propose that this combined observational‑modeling approach be adopted for future investigations of large continental earthquakes, especially those that generate significant off‑fault damage, as it can substantially narrow down plausible rupture scenarios and improve seismic hazard assessments.