Oral Presentation

The advent of plate-tectonic theory gave a new basis for understanding earthquakes along the San Andreas fault – in the new paradigm, it was a transform fault between two great plates. This gave a secure start to quantitative seismic hazard. Allen (1968) had suggested that only the 1906 and 1857 sections produced great earthquakes. If so, could we divide the SAF’s ~30 mm/yr geological slip rate into their slips to get the times between great earthquakes? In 1975 I started to pursue both terms in this equation: The numerator – slip during great earthquakes. And the denominator – the SAF’s millennial slip rate. Offsets measured after 1906 and offset landforms along the 1857 reach turned out to be complex, varying from ~3m to ~10m. But, larger offsets at any one site were often ~multiples of the smallest offset. Did these data imply that the slip rate also varied along fault and that historical ruptures faithfully repeated? Or was slip rate constant and recurrence variable along the fault? Millennial slip rates determined in the 1980s implied rates for the SAF system of ~34 mm/yr. Encouragingly, these gave a reliable background loading rate for repeating ruptures and for the intervening creeping reach. Early on, I also envisioned a suite of sites along the 1857 reach that would constrain maps of individual ruptures over the centuries. Pallett Creek was the first such paleoseismic site – ten ruptures in the past ~1200 years. But individual intervals between events varied from about a half to over two centuries. Moreover, 3-D excavations there demonstrated that the amount of slip varied greatly from one event to the next, so the hypothesis of uniform slip from event to event had to be discarded. The great variation in recurrence intervals and slips at Pallett Creek inspired new questions: Were San Andreas events modulated by slip on nearby faults? Or could the fundamental mechanics of even a perfectly rectangular, solitary fault produce such irregular slips and intervals? By 1989 I realized, with chagrin, that it would take me, my students and colleagues decades to uncover precise and abundant stratigraphic and geomorphological records adequate for our fault-mechanic colleagues to resolve the issue credibly. So, I looked for a great fault with a far simpler geometry and a more-easily recoverable stratigraphic/geomorphic record. That fault was the Sunda megathrust offshore Sumatra, and I began working on its coral reefs in 1991.