Advanced Technologies at the Northeast Fisheries Science Center
We use a variety of technologies within the center to advance our science.
Image
Advanced technology helps us collect better data about marine ecosystems. We use these tools to enhance traditional data collection, which leads to improved science for sustainable fisheries management and protected species conservation. At the Northeast Fisheries Science Center, we apply this technology across many different fields to achieve our mission. Below are a few examples of how we use advanced technology in our work.
Optics Technology
Optics technology helps advance and improve how we collect data to better understand marine species and ecosystems. Optics uses sophisticated underwater cameras to take images and videos that we use as data. Surveys using optics technology can collect large volumes of data quickly—something that wasn’t possible until relatively recently. We’re using optics technology to:
- Count and measure fish and invertebrates in our surveys that support stock and ecosystem assessments
- Identify protected whales, dolphins, seals, and sea turtles to inform stock assessments
- Characterize plankton communities that support marine ecosystems
- Monitor commercial fisheries catch for improved bycatch and discard estimates
One way we’ve been using optics technology is with our Atlantic Sea Scallop HabCam Survey. This survey uses sophisticated cameras to collect scallop abundance and distribution data.
Uncrewed Marine Systems
Uncrewed systems, such as autonomous underwater and surface vehicles, are increasingly becoming capable of conducting marine monitoring and research missions traditionally conducted using crewed systems. Using uncrewed systems allows us to address operational challenges for surveys, in particular at times or in regions that previously have not been possible for long-standing surveys. We’re using uncrewed systems to:
- Count and measure fish and plankton using uncrewed surface vehicles equipped with active acoustics
- Conducting optical Atlantic sea scallop and plankton optical surveys using autonomous underwater vehicles
- Using gliders to conduct passive acoustic surveys for fish and protected species
Active Acoustics
We have been collecting active acoustic data during fisheries-independent resource surveys and fisheries-dependent activities since 1998. Active acoustic technology monitors the entire water column, providing continuous data on the three-dimensional distribution of acoustically detectable marine life. We use active acoustic technology to
- Provide data for stock assessment
- Explore ecosystem services
- Conduct surveys using uncrewed marine systems in areas and at times that are unreachable by vessels
Passive Acoustics
Passive Acoustic Monitoring uses advanced technologies, including hydrophones (underwater microphones), to collect audio data in different regions of the ocean. We use hydrophones with specialized "listening ranges" to record species calls and environmental and human-made sounds that occur at different frequencies. Once we collect the audio data, we use acoustic detectors to help us find and measure the sounds within the recordings.
Image
The hydrophones are mounted on a range of platform types including being anchored to the seafloor, towed from the stern of research vessels, or mounted on autonomous ocean gliders and buoys which transmit data back to shore in near real-time. These technologies have advanced greatly in recent years and have allowed audio data to be gathered at times when visual surveys cannot operate (during bad weather or at night).
We now have long-term acoustic datasets spanning more than 20 years. This enables us to study changing trends in species occurrence and their soundscapes (acoustic environments). Our ability to identify when certain species are calling in a given area and to study their acoustic environment is critical for conservation management in near real-time and over decades.
‘Omics and Genetics
‘Omics
Genomic science examines an animal’s or plant’s DNA directly, which is changing the way we monitor marine ecosystems. While traditional genetics focuses on single genes, genomics allows us to study the entire genetic blueprint of organisms and their interactions with the environment. This advanced technology opens up new lines of evidence-based fishery assessment and management:
- Identifying species with DNA in the environment
- Estimating animal ages with markers that attach on the DNA
- Delimiting populations based on their genomic profile
- Using the DNA as a tag to estimate population size
Learn more about ‘omics in the Northeast
Satellite Technology
Satellite Oceanography
Image
Satellite remote sensors collect a variety of valuable oceanographic data, such as high-resolution ocean color imagery and sea surface temperatures. Scientists at NOAA Fisheries are using these data to assess ecosystem change, inform fisheries management, force ecosystem models, monitor environmental events like harmful algal blooms, and more.
- Looking Down To Improve The View From Above
- Dinoflagellate Bloom Dominates the Gulf of Maine
- Satellite Data
- Satellite Oceanography
Satellite Image Identification
Geospatial Artificial Intelligence for Animals
The Geospatial Artificial Intelligence for Animals initiative is advancing the use of very-high-resolution commercial satellite imagery and cloud-based geospatial workflows to support broad-area monitoring of whales and other marine species. Our current focus is on systematically collecting satellite imagery in priority habitats and building high-quality, expert-annotated training datasets that will underpin future automated detection capabilities. By establishing scalable infrastructure for ingesting, storing, and reviewing large volumes of imagery, we are laying the operational and scientific foundation for space-based marine monitoring to inform protected species management and ocean planning.
Images to Abundance
Image
NOAA Fisheries is launching a new national program focused on advancing the use of automatic animal detectors in still images. These images are captured during aerial abundance surveys, like those conducted in the AMAPPS project, for cetaceans, seals, sea turtles, and seabirds using the KAMERA 9-camera system. The program aims to create a fully automated and flexible workflow, encompassing the entire process:
- Data Acquisition: From the cameras in the aircraft to cloud image storage
- Processing and Annotation: Using Video and Image Analytics for Marine Environments for image annotation and testing detector models
- Machine Learning and Identification: Develop machine learning models to assist humans in detecting animals, identifying them to species, and generating output files
- Final Output: Producing files that can be used to estimate unbiased absolute abundance using spatial density modelling techniques
Satellite Tagging (Marine Mammals)
When It Comes to Tagging Seals, Teamwork Makes the Dream Work!
Short-Term Tagging Of Rare Whale Takes A Step Forward
Advances in Machine Learning Applications
Automated-image classification via machine learning is an emerging technology with the potential to improve video review, which would lower the costs of Electronic Monitoring (EM) programs. The NEFSC is collecting fish images from its biannual bottom trawl survey to build a groundfish image library. The image library will serve as a training tool for EM machine learning applications that could result in data analysis efficiencies and lower program costs for the fishing industry.
From this work, we will determine if this technology can estimate fish size and identify fish species to the level needed by managers and scientists. Results from this project will develop recommendations for how to move this technology from the scientific survey setting to fishing boats. The goal is to develop an algorithm for annotating EM footage in open source software products for operational programs. In EM programs, these tools could potentially collect species and weight information as the fishing crew handles the catch under a camera's view before discarding or assist with monitoring adherence to catch retention requirements. In addition, this technology could reduce the amount of video collected by using activity recognition tools (e.g., detecting crew on deck). Incorporating machine learning applications would increase catch reporting accuracy while expanding the use of EM to monitor fisheries.
Modeling
Multispecies and Ecosystem Modeling for the Northeast Shelf Ecosystem