Earth Institute Research Projects

Collaborative Research: Persistent Presence in the Ocean Interior: Developing a Low-power, Autonomous System for Geo-referenced Navigation

Lead PI: Dr. Andreas Thurnherr

Unit Affiliation: Ocean & Climate Physics, Lamont-Doherty Earth Observatory (LDEO)

January 2017 - December 2019
Project Type: Research

DESCRIPTION: Facilitating an accurately navigated persistent presence for the interior of the deep ocean has the potential to transform how oceanography is conducted. This capability is currently not provided by existing technologies; however, if developed would transform the scope of projects undertaken by underwater vehicles with longer range and endurance than can currently be deployed unsupported from research ships. This research develops a new low power navigation system that improves navigational accuracy from 1000s of meters to 100s meters. This advance will enable multiple new lines of oceanographic investigation. Examples include deployment of coordinated glider fleets to investigate complex physical-biogeochemical interactions; deep studies of topographically induced mixing; long-range characterization of seafloor habitats/ecosystems at the scale of entire ocean basins; and better resolution of the scales of bottom-boundary-layer processes in regions with steep underwater terrain. The researchers will engage in outreach activities including undergraduate participation in our research and continuation of established collaborations with local high school robotics and environmental science classes.  This research develops and tests a low-power acoustic positioning system that enables accurate externally aided navigation in the deep ocean. A glider (or fleet of gliders) operates at depth while an autonomous surface robot (ASV) follows on the sea surface. The ASV transmits its geo-reference position to the gliders at a regular interval. Each glider then independently employs a precision time base (provided by chip-scale atomic clocks) and array processing to determine its position relative to the ASV. Each ASV combines this relative position estimate with the ASV's position encoded in the received packet to compute its geo-referenced position. System performance depends on a number of factors including the precision of vehicle attitude sensors. Accuracies of 250-400m are possible at 5000m depth using low-power sensors already in use on the glider. Hence, for deep-diving gliders, this method will provide a 10-100 fold improvement in positioning accuracy over the current paradigm (i.e., infrequent GPS fixes) and would allow vehicles to spend more time at depth making relevant observations. This research will enable new operating paradigms, advance observational capabilities and facilitate spatially denser observations than were previously possible. Furthermore, the new system will also improve the ability to estimate depth-averaged ocean currents. The developed capability is also a prerequisite for coordinated motion control of multiple deep-diving AUGs, further increasing the density and cadence of deep ocean observations, providing an ever closer-to-synoptic view of the deep ocean interior. The developed system will be tested first in the tank and then in two ocean cruises including a Year 3 deployment on a Seaglider.