The relation between short- and long-term deformation in actively deforming plate boundary zones.

Satellite-based measuring systems are making it possible to monitor deformation of the Earth's surface at a high spatial resolution over periods of several decades and a significant fraction of the seismic cycle. It is widely assumed that this short-term deformation directly reflects the long-term pattern of crustal deformation, although modified in detail by local elastic effects related to locking on individual faults. This way, short-term deformation is often jointly inverted with long-term estimates of fault slip rates, or even stress, over periods of 10 s to 100 s kyrs. Here, I examine the relation between these two timescales of deformation for subduction, continental shortening and rifting tectonic settings, with examples from the active New Zealand and Central Andean plate boundary zone. I show that the relation is inherently non-unique, and simple models of locking on a deep-seated megathrust or decollement, or mantle flow, provide excellent fits to the short-term observations without requiring any information about the geometry and rate of surface faulting. The short-term deformation, in these settings at least, cannot be used to determine the behaviour of individual faults, but instead places constraints on the forces that drive deformation. Thus, there is a fundamental difference between the stress loading and stress relief parts of the earthquake cycle, with failure determined by dynamical rather than kinematic constraints; the same stress loading can give rise to widely different modes of long-term deformation, depending on the strength and rheology of the deforming zone, and the role of gravitational stresses. The process of slip on networks of active faults may have an intermediate timescale of kyrs to 10 s kyrs, where individual faults fail piecemeal without any characteristic behaviour. Physics-based dynamical models of short-term deformation may be the best way to make full use of the increasing quality of this type of data in the future. This article is part of a discussion meeting issue 'Understanding earthquakes using the geological record'.