How Well do the Mentawai Corals Constrain the 1797 Tsunami?

Vertical coseismic displacements recorded in corals on the Sumatran forearc offer a unique opportunity to reconstruct past earthquakes on the Sunda megathrust. These observations not only constrain the time and magnitude of the 1797 event but also place some bounds on possible slip distributions for this great earthquake. Here we explore the range of slip distributions which are consistent with the coral data. We choose a brute-force technique in which the displacements at the corals are computed for many possible, stochastic slip distributions. The technique makes no assumptions about the earthquake other than the geometry of the fault and the autocorrelation function of the slip distribution which limits the post-seismic strain on the megathrust. We expose several internal inconsistencies in the data with respect to a double couple event on a single inclined plane. We also show that while the along strike length of the rupture is reasonably well constrained, the down-dip distribution of slip, a first order control on the tsunamigenic potential of the earthquake, is only poorly resolved. Numerical simulations of tsunamis that would have resulted from possible 1797 earthquakes may help to advise on possible hazard from a near future Mentawai earthquake.