Our study integrates field geologic observations, geodetic measurements, and geophysical modeling to provide a detailed anatomy of a great megathrust rupture. Some of the interest in the paper stems from a recent flurry of work along the Sunda subduction zone, the source of the tsunamigenic 2004 M 9.2 and destructive 2005 M 8.7 ruptures.

“Some of the interest in the paper stems from a recent flurry of work along the Sunda subduction zone, the source of the tsunamigenic 2004 M 9.2 and destructive 2005 M 8.7 ruptures.”

Because it addresses a large subduction zone earthquake in detail, the paper is also relevant to subduction zones worldwide as well as to more general earthquake geology and seismotectonic studies.

There are several overlapping components of the study—geology, geodesy, and geophysical modeling—and, when combined, they provide one of the most detailed maps of surface deformation and fault slip obtained for a megathrust rupture.

Each component of the study required new or refined techniques. For example, coral heads have previously been used by Fred Taylor of the Institute for Geophysics, University of Texas, Austin, and others for determining coseismic uplift, but we expanded the technique to measure large amounts of uplift and even subsidence by using a numerical tide model. We also used remote sensing imagery to constrain uplift in places we couldn't reach in the field, a technique developed by Aron Meltzner of the California Institute of Technology.

The slip model developed by Ya-ju Hsu and Mark Simons of the California Institute of Technology’s new Tectonics Observatory is unique in several ways, including their joint inversion of the coral and GPS data that honors the extensive spatial distribution but higher uncertainty of the corals.

This paper provides a snapshot of how rupture occurred on a deeply buried fault, or megathrust, along a subduction zone where two of Earth’s plates collide. We made measurements of surface movement using very low-tech but widespread natural tide gauges—large coral heads—and a few high-tech Global Positioning System (GPS) stations, several of which were installed mere weeks before the rupture. We were lucky because islands near the plate boundary allowed us to precisely measure how the surface rose or sank, and we were able to use our measurements of ground motion to model how slip occurred deep in the earth. This is often difficult to do in subduction zones because the slip during megathrust earthquakes typically occurs deep beneath the ocean.

One of our major findings was that the earth rose nearly three meters in some places, and sank well over a meter in others. We mapped the line of zero change in between the zones of uplift and subsidence which is important for determining where slip occurred on the underlying fault.

We found that the bulk of slip in 2005 took place beneath the islands rather than beneath the sea—this probably helps explain why the 2005 tsunami was smaller than might otherwise have been expected—and that structural boundaries along the subduction zone appear to have influenced how slip occurred.

This study was a collaboration between geologists, geodesists, and geophysical modelers in both the US and Indonesia, and the paper relied very much on the contributions of each co-author. Professor Kerry Sieh, of the Division of Geology and Planetary Sciences at Caltech, along with multiple students and colleagues, has spent over a decade characterizing the earthquake history and hazards of the Sunda subduction zone and his scientific work and interest in earthquake hazards in the region laid the groundwork for our rapid response. I came to Caltech to collaborate with Kerry in Sumatra but neither of us knew then that the 2004/2005 ruptures would form the basis of our work together.

Because the fieldwork began closely on the heels of the March 28, 2005 rupture, we encountered the usual challenges of working in a disaster area, but any setbacks we faced paled in comparison to the obstacles (many still ongoing) faced by the residents of the affected islands.

Richard Stone has written a very nice article for Science magazine that highlights the long-term effects of coseismic uplift and subsidence on the people living along the coasts of Nias, Simeulue, and the Banyak islands.

In the long term, the surface motions that occurred in 2005 will reverse themselves as the subduction zone recovers elastic strain, but over the lifetime of most of the residents alive today, the uplift and subsidence are permanent. This means that the loss of reefs, or the flooding of houses, will extract a toll on many communities long after the shaking and tsunamis are gone.

Governments and aid agencies can easily appreciate the destruction caused by groundshaking or tsunamis associated with megathrust rupture, but we don’t yet fully appreciate the more subtle but often substantial long-term damage that accompanies coastal uplift or subsidence. Unfortunately, this is being replayed in the Solomon Islands following the April 1, 2007 M 8.1 rupture, where uplift of several meters and subsidence of nearly a meter is having serious local impacts.