Geological samples are collected by lowering a steel pipe called a core barrel of 9.5m long inside a drill pipe. Supposing a core is recovered of a length of 100m, then a length of 9.5m is drilled at a time, after which the core barrel is pulled up. This is then repeated more than ten times.

The fault zones that qualified on the basis of the LWD data are found at two locations of 720m and 820m below the seafloor. It takes a tremendous amount of time to collect all geographical samples up to those depths by repeatedly bringing up and lowering the core barrel.

It was therefore decided to employ a strategy of drilling to the target depth without collecting geological samples on the way. Keeping an eye on the weather forecast, and after calculating the remaining time on the basis of progress made, the target fault zone was aimed at while skipping stretches as appropriate.

　Upon hearing that they had hit a hard layer after reaching a certain depth, Saito recalls thinking, “This is it!” He had in fact postulated, based on the LWD data, that there was a hard layer right above the clay layer.

　Saito, deducing that the drill bit had arrived at the clay layer, then made a proposal. He suggested drilling in increments of several meters in order to raise the collection rate. Sample collection rates in drilling fault zones often fall below 40%. This is the same whether drilling is at 9.5m or 2m intervals. So drilling at small increments increases the possibility that samples are recovered from the target fault zones. It goes without saying however that the method of drilling at several meter intervals takes time. Says Ujiie was our definite desire to collect geographical samples from the fault that pushed the decision. The decision was taken to drill at increments of several meters.

　May 21st, and the 17th core was lifted on board. Seeing the sample, the scientists shouted in delight. Ujiie, who saw it, tells, “The rock was finely sheared. It was clear at the first glance that this was the fault zone.” The layer above the fault zone consists of sand and fine particles which have been deposited and have solidified. When a layer slips as a fault at the time of an earthquake, the solidified layer fractures and is finely sheared. This sample presented the same picture as fault zones which can be seen in layers on the surface.

　“This was the time we had been waiting for. We were excited that we had been able to recover a sample through water of such a great depth, and moreover from such a deep subseafloor layer, and that this was our first one from the plate boundary, our target,” tells Mori.

　Since then, it was decided make another attempt at installing temperature sensors, the other JFAST objective, in July. All people involved with JFAST strongly felt that they couldn’t miss capturing frictional heat data on the ground, in order to gain an understanding of the massive earthquake.