The Nankai Trough (seen in the inset) is a very well studied region. A number of people in this room have written important papers about this region. It involves the subduction of the Philippine Sea plate underneath southern Japan at a rate that is known to be at about 4 centimeter a year. And it is the site of many many historic earthquakes. The most recent large ones have been the Nankaido earthquake of 1946, which probably ruptured this entire area, and an earthquake two years earlier that ruptured this area here.
In cross-section (fig.40) you see the goal of the seismic experiment is to image this region, to go from the part of the subduction zone that is aseismic down into the part that ruptured in 1946, and to try to understand the variation in properties that one sees.
Of course, seismology, while it's capable of producing wonderful images, is not the same as ground truth. And what you see here is one existing ocean drilling hole some proposed shallow holes seaward of the trench. The notion that from a riser drill ship one could from this position drill some ten kilometers deep, and actually drill through the seismic part of this plate interface, is a very exciting potential possibility for testing the results of this seismic survey.
Let me mention a second seismic experiment that is going on off Japan (fig.41), because some colleagues of mine are involved. This is an experiment that is jointly being carried out by the University of Tokyo Ocean Research Institute and the Carnegie Institution of Washington to instrument two ODP holes, (Leg.186 to those who follow it). We'll drill these two holes here, which are above the seismic zone in the Japan Trench, where the Pacific plate is subducting at something like ten centimeters a year beneath eastern Japan. The holes will not penetrate down to the seismic zone, (the current drill-ship does not have that capability)but even above the seismic zone these two holes (fig.42) will be very important for understanding both the seismicity and the long-term deformation of this particular system.
I want to mention one particular instrument that will be part of a package, partly because it's a Carnegie instrument, but partly because it records a different kind of data from a conventional seismometer. The instrument package will contain seismometers but also in the instrument package (fig.43) is something called a strainmeter, which records very sensitive changes in volume and strain over a very broad frequency range. Now why is that important ? In this part of the Japan Trench, we think we know how fast it is subducting, and we think we know how many earthquakes that should produce in a given time if all of the slip on the fault generated large earthquakes. And you can calculate the slip from the known earthquakes, and that amounts to only about a quarter of the expected slip. So we have several possible explanations for what's going on. One of them is that some very large earthquakes are not part of our catalogue - they haven't happened yet. But another possibility is that some of the slip is going on by other means not involving earthquakes.
And I show you this example (fig.44) because although an old example it is one of the first clear examples of a kind of event that is different from a normal earthquake, it is called a slow earthquake. There was a normal earthquake that occurred here in this region between the Izu Peninsula and Oshima. But it was followed immediately by another event in this region that was very different. And here are three strainmeter records that show that difference - the normal part of the earthquake is shown as the steps in strain, and then over the succeeding hour or more, the strain continues to change. In fact the change in strain was comparable over that hour to what happened during the normal part of the earthquake. And my colleagues Selwyn Sacks and Alan Linde and their Japanese colleagues called this part of the event a slow earthquake. These kinds of events have shown up elsewhere, although the instrumentation is not sufficiently widespread to record them and characterize them, but they're thought to be fairly common on different kinds of plate boundaries.
