Impact of employing a waveglider on GNSS-Acoustic survey along the Japan trench

Following the 2011 Tohoku Earthquake, we constructed seafloor geodetic benchmarks for GNSS-Acoustic measurement at twenty sites along the Japan trench in September 2012 and have started repeating surveys since then. Reliable horizontal displacement rates were obtained to date for a sufficiently long period of surveys, which revealed the coexistence of viscoelastic relaxation and after slips from place to place. Further analysis to estimate vertical motion, we preliminary exposed regions of uplift and subsidence, although the expected errors were still significant. However, the pattern of vertical motion gives independent information from the horizontal ones for verifying viscoelastic models and evaluating the extent of after slips.

We introduced an unmanned autonomous vehicle, called Waveglider (WG), as a surface platform instead of a ship, which overcomes the deficiency in ship-time in the sense of budget and human resources. Actually, data obtained by WG bear comparison with that by shipboard survey, and even the onboard-processed data can be transmitted to the onshore station via satellite system nearly in realtime. Moreover, significantly intensive use of WG helps increase the survey frequency, which can have a chance to identify slow slip events; for instance, marine seismometers have revealed the existence of slow events off the Sanriku region near our survey sites. 

The WG deficit is sometimes trapped against sea current even along the Japan trench, typically when over 2 knots, and fails into low power conditions depending on weather, season, and solar culmination altitude. Well-organized planning and operation may reduce such deficit. More practically, the slower speed of WG prevents efficient moving survey over a transponder array, especially for deep (>5000m) sites having a large footprint of the array, which is typical for our sites. Insufficient moving survey degrades the accuracy in vertical crustal movement, the importance of which increases to monitor the afterslip distribution as noted above. Then we solve this problem using a different approach that utilizes different incident angles even in point survey by employing two concentric triangles of different sizes for a six transponder site or a triangle with a centered one for a four transponder site.

If WG operation would become more common and can be appropriative, a fully continuous survey will be realized at a specific site without a moored buoy. This will be valuable not only for detecting temporal phenomena like slow slip events but also for disaster mitigation to monitor offshore fault failure in realtime. In addition, such prompt measurement just after a large earthquake reveals rapid postseismic deformation in an early stage. Long-term continuous operation requires particular battery specifications and operational fashion for seafloor transponders. We are also designing such transponders at this moment.