Jumping Earthquakes: limits to rupture forecasting exposed by instantaneously triggered earthquake doublet

An important question in understanding the potential magnitude of earthquakes, and consequently the hazard certain faults may pose, is the distance over which earthquakes can jump during rupture. This is because a critical observation of earthquake scaling which holds true is that the longer a fault rupture, the larger the earthquake. In a bigger earthquake, the ground shaking is more severe, it occurs over a wider area and lasts longer; exposing more buildings and people to a greater level of hazard.

The previous consensus on the control of the maximum likely leap in an earthquake was that an offset between faults of 5 km would probably be enough to stop a rupture – the gap being too much of a physical barrier for the earthquake to jump across. This limit of 5 km is used in some standard seismic hazard assessments.

However, an international team of researchers, including current and ex- members of the Oxford Earth Sciences department, have forensically unpicked from a seismic event that struck Pakistan in 1997, a pair of large earthquakes that had been previously catalogued as one. In a study published this week in Nature Geoscience, they combine satellite observations with seismology, to show that this pair of earthquakes involved a massive jump of 50 km between fault segments during the rupture – this is 10 times larger than the current accepted rule that is used in assessing earthquake hazard.

Satellite radar image of the ground deformation resulting from the 1997 Pakistan earthquake. The two bullseyes of deformation in this image (centre and lower right) were crucial in revealing the doublet nature of this earthquake and the large distance over which the rupture jumped.

This finding will fundamentally change the way in which earthquake hazard scenarios are run in regions of distributed faulting, because the network of faults which could potentially be triggered as the rupture proceeds would be greatly expanded. Therefore, not only is it crucial to determine the range of jumps feasible in earthquakes, such as in this study, it is also critical to establish the distribution of faults within the crust which could be potentially triggered. This will be no easy task, given that many faults, such as the pair involved here, are hidden underground and lack a clear surface expression to give away their presence. There are large regions of the world, some with huge cities, which have these buried faults beneath them – places such as Southern California, Iran and east China. The goal of identifying active faults is one of the major research avenues for the Earthquake Geology and Geodesy group within the department.

This study is published in the journal Nature Geoscience and involved current members of the department John Elliott and Barry Parsons, as well as past members of the group – Ed Nissen, Al Sloan, Gareth Funning and Tim Wright.

The work was funded in part by the Natural Environmental Research Council (NERC) through the Earthquake without Frontiers (EwF), Looking Inside the Continents from Space (LiCS) and the Centre for the Observation and Modelling of Earthquakes, Volcanoes & Tectonics (COMET).

Research Paper: Nissen, E. K., Elliott, J. R., R. A. Sloan, T. J. Craig, G. J. Funning, A. Hutko, B. E. Parsons & T. J. Wright (2016) Dynamic triggering of an earthquake doublet exposes limitations to rupture forecasting, Nature Geoscience, doi:10.1038/NGEO2653

Fold and thrust belt of the western Sulaiman mountains of Pakistan. Hidden beneath these folds is the pair of large fault planes that ruptured in the 1997 Harnai earthquake, jumping 50 km between segments. Image: Landsat 8 false colour image (742).