Deep-currents along a rift valley of Central Indian Ridge observed by AUV "r2D4"

In December 2006, the Institute of Industrial Science made geological and chemical investigation along rift valleys off the Rodriguez Island in the Central Indian Ridge with an AUV "gr2D4"h and found hydrothermal activities in one of the valleys, named the Great Dodo Lava Plain (water depth of approximately 2700m). In this present study, based on the AUV"fs navigation data taken from the investigation, we figured out the distribution of deep-currents in the Great Dodo Lava Plain and analyzed those currents and their relations with the tidal current and water-mass property. For current velocity, we used water-tracking velocity data measured by the Doppler Velocity Log (DVL) at a layer under the AUV. When the AUV was close to the sea bed, the velocity was adjusted to the absolute current velocity with bottom-tracking velocity measured by the DVL. When the bottom-tracking velocity was not available, the velocity was adjusted with AUV"fs velocity measured by the Inertial Navigation System. The adjusted current velocity data in the Great Dodo Lava Plain were observed for 6.5-hours in the range of 18.32°S-18.45°S, 65.28°E-65.37°E, and most of them reached 20 cm s-1 or more. The direction of the currents is between northwest and northeast, which almost corresponds to the direction of the valley"fs extension (north-northwest). The observed current speed is much higher than the tidal current speed predicted from a barotropic tidal model (< 3 cm s-1), but its meridional velocity component is stronger than the zonal component as well as the observed velocity. The predicted tidal current velocity shows semi-diurnal period, and that is also the case in the observed meridional velocity component. Thus, it is expected that the deep-currents and tidal currents are influenced by the bottom topography extending north-northwestward. For error consideration, we compared near-surface current velocities measured by the AUV and by the shipboard acoustic Doppler velocity profiler of the parent vessel. Then, the directions roughly correspond although the current speed measured by the AUV is rather low. This might be because bottom-tracking data measured by the DVL were not available at near-surface. In our presentation, we will also show the analysis in terms of water-mass properties and the other results from the data measured in the Mariana Trough in May 2004 and in the Izu-Ogasawara Island Arc in August 2005.