Machine Learning-Based Automated Detection of Seafloor Gas Seeps

Past Expedition

Overview

Recognizing the importance of seeps and the significant potential for further seep discovery, a research team has been developing an efficient and cost-effective machine learning-based software system to automatically detect seafloor gas seeps (i.e., cold seeps) in mapping sonar water column data. Seeps are hotspots of deep-sea biodiversity, an important part of the global carbon cycle, indicators of potential marine geohazards, and an energy production resource. However, the deep ocean is largely unexplored, and most seeps are discovered through manual visual analysis of water column sonar imagery, which is time consuming, costly, and inconsistent. This project is designed to automate that process, accelerating the speed, accuracy, and consistency of seep discovery, and resulting in improved understanding of seep abundance and distribution.

Training the Model

The core element of the new system is a machine learning model, which needed to be trained to detect seeps. To train and validate the model, the team relied on multibeam sonar water column data collected by NOAA Ocean Exploration during expeditions along the U.S. Atlantic margin on NOAA Ship Okeanos Explorer. They visually reviewed over 428,396 multibeam sonar water column images — searching for gas bubble plumes — and digitally labeled them, noting the presence or absence of seep targets. Seep targets were present in 6,583 (1.54%) of these images. Where seeps were present, the researchers drew bounding boxes around the seep targets to indicate their locations, which resulted in 8,324 bounding boxes.

Seafloor gas seeps are detected through identification of associated bubble plumes in multibeam echosounder sonar imagery. These plumes generally appear as near vertical lines of elevated acoustic reflectivity. Imaged bubble plumes are traced to their point of intersection with the seafloor to establish the position of the source gas seeps. *Image courtesy of Machine Learning-Based Automated Detection of Seafloor Gas Seeps.*
Download largest version (jpg, 1.18 MB).

This map shows seafloor gas seeps (red points) and the total sonar survey area (gray) on the northern U.S. Atlantic margin used by the research team to train and validate the machine learning-based automated seep detection model. *Image courtesy of Machine Learning-Based Automated Detection of Seafloor Gas Seeps.*
Download largest version (jpg, 3.15 MB).

The research team used images of the water column produced by multibeam sonar systems to search for gas seep bubble plumes, like the one shown here (inside white box) rising from the seafloor (red line). *Image courtesy of Machine Learning-Based Automated Detection of Seafloor Gas Seeps.*
Download largest version (jpg, 1.49 MB).

Next Steps

Next steps for this project include applying the approach to multibeam sonar data at the National Centers for Environmental Information that have not yet been reviewed for the presence of seeps, developing algorithms to estimate seafloor seep origin (latitude, longitude, and depth) for each gas plume the model detects, and continuing refinements. Additionally, the team is working toward using the system while multibeam sonar data are being collected to detect seeps in near real time. This will give researchers the ability to rapidly recognize and potentially respond to the presence of gas seeps while the survey ship is at sea.

During a remotely operated vehicle dive offshore Virginia north of Washington Canyon, methane gas bubbles flow in small streams out of seafloor sediment. Quill worms, anemones, and patches of microbial mat can be seen in and along the periphery of the seepage area. Sonar images of bubble plumes like these were used to train the machine learning-based automated seep detection model. *Image courtesy of NOAA Ocean Exploration, 2013 ROV Shakedown and Field Trials in the U.S. Atlantic Canyons.*
Download largest version (jpg, 1.6 MB).

Expected Outcomes

The primary outcome of this project is a new technology that is expected to increase the speed, accuracy, and consistency of seep detection, while also decreasing costs and personnel requirements. Another anticipated outcome is the discovery of new seeps. Together, these and other outcomes will improve our basic understanding of the quantity and location of seafloor gas seeps around the world, providing critical deep-ocean data necessary for effective management and protection of ocean resources.