New AI Tool Revolutionizes Iceberg Detection in the Southern Ocean
In a groundbreaking study published in the journal Remote Sensing of the Environment, researchers have unveiled a cutting-edge AI tool designed to detect icebergs in the Southern Ocean. This innovation marks a significant leap forward in the ability of scientists to monitor the complete life cycle of icebergs across Antarctica using satellite data.
Revolutionizing Iceberg monitoring
Icebergs, pivotal in ocean dynamics, influence freshwater release, nutrient distribution, and impact primary productivity. The hazards they pose to shipping demand accurate real-time information on their location and size.
The newly developed AI tool harnesses data from Synthetic Aperture Radar (SAR), part of the Sentinel-1 satellites, enabling the identification of icebergs even in regions densely covered with sea ice—a capability unprecedented until now.
Operating on the principle that icebergs reflect microwaves due to their crystalline surface structure, the AI tool utilizes SAR’s microwave signals to create detailed images. This method ensures data collection day or night, irrespective of cloud cover—a common occurrence over the Southern Ocean.
The study, funded by The Alan Turing Institute, successfully demonstrated the algorithm’s prowess over a 12-month period, identifying nearly 30,000 icebergs, most measuring 1 km² or less.
Researchers focused their study on the Amundsen Sea Embayment in West Antarctica, near the calving front of the Thwaites Glacier. This site’s diverse terrain, comprising open water, sea ice, and a concentration of icebergs, provided an ideal testing ground for the AI tool.
Ben Evans, lead author and member of the British Antarctic Survey (BAS) AI Lab, emphasizes the tool’s accuracy and efficiency, comparable to alternative methods but surpassing many without requiring human input.
AI’s crucial role in climate change monitoring
Calving from the Antarctic Ice Sheet into the Southern Ocean serves as a major contributor to ice loss, potentially signaling increased sea level rise. The AI tool’s real-time monitoring capability positions it as a valuable asset in identifying changes in iceberg numbers, size, and pathways—key indicators of climate change impact. The team is currently scrutinizing data since the inception of the Sentinel-1 mission in 2014, aiming to unveil insights into Antarctica’s response to climate change.
Scott Hosking, Head of the BAS AI Lab and Co-Director for the Turing Research and Innovation Cluster in Digital Twins at The Alan Turing Institute, highlights the broader vision of developing a digital twin of Antarctica.
This initiative aims to integrate and share data across various polar infrastructure and tools, ranging from automated underwater vehicles to AI models. The overarching goal is to enhance decision-making and maintain the UK’s forefront position in polar science.
Read More
New AI Tool Revolutionizes Iceberg Detection in the Southern Ocean
In a groundbreaking study published in the journal Remote Sensing of the Environment, researchers have unveiled a cutting-edge AI tool designed to detect icebergs in the Southern Ocean. This innovation marks a significant leap forward in the ability of scientists to monitor the complete life cycle of icebergs across Antarctica using satellite data.
Revolutionizing Iceberg monitoring
Icebergs, pivotal in ocean dynamics, influence freshwater release, nutrient distribution, and impact primary productivity. The hazards they pose to shipping demand accurate real-time information on their location and size.
The newly developed AI tool harnesses data from Synthetic Aperture Radar (SAR), part of the Sentinel-1 satellites, enabling the identification of icebergs even in regions densely covered with sea ice—a capability unprecedented until now.
Operating on the principle that icebergs reflect microwaves due to their crystalline surface structure, the AI tool utilizes SAR’s microwave signals to create detailed images. This method ensures data collection day or night, irrespective of cloud cover—a common occurrence over the Southern Ocean.
The study, funded by The Alan Turing Institute, successfully demonstrated the algorithm’s prowess over a 12-month period, identifying nearly 30,000 icebergs, most measuring 1 km² or less.
Researchers focused their study on the Amundsen Sea Embayment in West Antarctica, near the calving front of the Thwaites Glacier. This site’s diverse terrain, comprising open water, sea ice, and a concentration of icebergs, provided an ideal testing ground for the AI tool.
Ben Evans, lead author and member of the British Antarctic Survey (BAS) AI Lab, emphasizes the tool’s accuracy and efficiency, comparable to alternative methods but surpassing many without requiring human input.
AI’s crucial role in climate change monitoring
Calving from the Antarctic Ice Sheet into the Southern Ocean serves as a major contributor to ice loss, potentially signaling increased sea level rise. The AI tool’s real-time monitoring capability positions it as a valuable asset in identifying changes in iceberg numbers, size, and pathways—key indicators of climate change impact. The team is currently scrutinizing data since the inception of the Sentinel-1 mission in 2014, aiming to unveil insights into Antarctica’s response to climate change.
Scott Hosking, Head of the BAS AI Lab and Co-Director for the Turing Research and Innovation Cluster in Digital Twins at The Alan Turing Institute, highlights the broader vision of developing a digital twin of Antarctica.
This initiative aims to integrate and share data across various polar infrastructure and tools, ranging from automated underwater vehicles to AI models. The overarching goal is to enhance decision-making and maintain the UK’s forefront position in polar science.
Read More