Autonomous Localization of Seafloor Fluid Flow Sources

Past Expedition

Primary Goal

Develop a fully autonomous algorithm enabling autonomous underwater vehicles to trace biogeochemistry signatures in the water column that are indicative of hydrothermal vents and cold seeps using real-time data from onboard sensors.

Dates

September 2024

Location

Pacific Ocean

Vessel

Exploration Vessel Nautilus

Primary Technology

Autonomous underwater vehicle Sentry, plume-tracing algorithm

Overview

To streamline the process of locating seafloor fluid flow sites such as hydrothermal vents and cold seeps, this research team is developing a fully autonomous algorithm to enable autonomous underwater vehicles (AUVs) to trace biogeochemistry signatures in the water column that are indicative of vents and seeps using real-time data recorded by onboard sensors, independent of human operators. In September 2024, during a NOAA Ocean Exploration Cooperative Institute expedition on Ocean Exploration Trust’s Exploration Vessel Nautilus, they tested a prototype of their plume-tracing algorithm during two AUV Sentry dives at Vailulu’u Seamount in National Marine Sanctuary of American Samoa.

An underwater vent — Hydrothermal vent on Vailulu’u Seamount in National Marine Sanctuary of American Samoa visually identified by remotely operated vehicle *Hercules* at the beginning of the expedition. This vent was the target site for the testing of the prototype of the plume-tracing algorithm developed for the Autonomous Localization of Seafloor Fluid Flow Sources project. *Image courtesy of Ocean Exploration Trust/NOAA Ocean Exploration.*
Download largest version (png, 2.72 MB).

The team selected Vailulu’u Seamount for this test because of its known hydrothermal activity. Early in the expedition, remotely operated vehicle Hercules revisited a hydrothermal vent on the eastern rim of the volcano’s crater, first discovered in 2019, and confirmed that it was still active.

A bathymetric map of Vailulu’u Seamount. The black dot marks the location of the hydrothermal vent on the eastern rim of the volcano’s crater that was visually identified by remotely operated vehicle *Hercules*. *Image courtesy of Autonomous Localization of Seafloor Fluid Flow Sources. Based on data from the 2017 expedition to Vailulu’u seamount (Herrera, S. et al. 2023).*
Download largest version (png, 153 KB).

To conduct the tests, Sentry, equipped with sensors for detecting anomalies in the water column associated with hydrothermal discharge, was deployed in the area of the vent site. While Sentry followed its pre-programmed route to conduct water column surveys, flying several tens of meters above the seafloor, sensor data were transmitted to shore in near real time for analysis by the plume-tracing algorithm. The algorithm processed the data collected over these initial large-scale surveys, identifying potential vent source locations and determining coordinates for the centers and boundaries of small-scale follow-up surveys to target the most probable source areas. These coordinates were then communicated back to the ship, where the AUV navigators redirected Sentry, and the exploration continued.

To view this video please enable JavaScript, and consider upgrading to a web browser that supports HTML5 video

Deployment of autonomous underwater vehicle Sentry from Exploration Vessel Nautilus during the E Mamana Ou Gataifale II — American Samoa expedition. Video courtesy of Autonomous Localization of Seafloor Fluid Flow Sources. Download largest version (mp4, 16.7 GB)

Outcomes

Evaluation of the algorithm’s performance suggests it performed as anticipated. During both dives, the algorithm successfully determined coordinates for follow-up surveys of areas most likely to encompass hydrothermal vent sources. These coordinates aligned well with those the dive watchstanders, who have extensive experience in hydrothermal plume tracing, would have selected in a traditional human-guided mission. The testing also confirmed that sudden decreases in oxidation-reduction potential (ORP) were the most sensitive and reliable indicators of hydrothermal activity.

There was another important outcome that has implications for the future use of uncrewed marine systems (on Earth and beyond): This project demonstrated, for the first time, that it’s possible to transfer necessary data and contextual information from an AUV in the deep ocean all the way to a remote center ashore and for guidance to then be sent back to the AUV, all while the vehicle is still submerged and continuing operations.

A graph showing time derivative of oxidation-reduction potential — Time derivative of oxidation-reduction potential (dORPdt) derived from the near real-time *in-situ* sensor data collected by autonomous underwater vehicle *Sentry* during the two surveys over the eastern rim of the crater of Vailulu’u Seamount on September 17, 2024 (the broadly spaced outer lines show data from the initial large-scale survey; the tightly packed lines in the center show data from the small-scale survey directed by the algorithm). Significant negative anomalies in dORPdt are indicated by the darkest dots and suggest where the source of the hydrothermal signals likely originates. The magenta dot marks the location of a known hydrothermal vent. *Image courtesy of Autonomous Localization of Seafloor Fluid Flow Sources.*
Download largest version (png, 93 KB).

A graph showing oxidation-reduction potential (ORP) — Screenshot of oxidation-reduction potential (ORP) from real-time *in-situ* sensor data collected by autonomous underwater vehicle (AUV) *Sentry* on September 17, 2024, displayed in the dashboard software used for AUV dive watchstanding. The left panel presents the time series of ORP during the initial survey, with a significant drop after 12:30 corresponding to the moment the vehicle intercepted hydrothermal anomalies. The right panel illustrates a planar view of the survey trajectories, where darker colors indicate large drops in ORP (see color scale with y-axis). *Image courtesy of Autonomous Localization of Seafloor Fluid Flow Sources. Dashboard software developed by Victoria Preston, Olin College of Engineering.*
Download largest version (png, 521 KB).

Next Steps

The remainder of the project will focus on further developing the autonomous plume-tracing algorithm and integrating it into Sentry’s command and control system to achieve fully autonomous plume tracing.

The surveys conducted in September 2024 showed that parameters used as indicators of hydrothermal activity, such as ORP, can vary significantly with depth. Additionally, hydrothermal discharge expands and disperses as it rises from the seafloor. To address these variabilities, future testing will explore the use of multilevel surveys at different depths to improve the precision of hydrothermal source localization.

This project is anticipated to improve the efficiency of AUV-based exploration and enhance the ability of scientists to conduct other operations at the same time while at sea. Ultimately, it has the potential to improve our knowledge about seafloor vents and seeps and the marine life they support — in our ocean as well as those on other ocean worlds.

Hydrothermal Vents and Volcanoes

Underwater Robots

Cold Seeps

Autonomous Localization of Seafloor Fluid Flow Sources

Overview

Outcomes

Next Steps

Education

Resources & Contacts