[ad_1]
For generations, the Arctic people have relied on seasonal sea ice, which grows and recedes throughout the year. Polar bears and marine mammals rely on it as a place to hunt and rest; Indigenous people fish from holes in the ice, known as wormwood, and use well-known routes through the ice to travel from place to place. But Arctic air and water are warming three times faster than the rest of the planet since 1971, according to a May 2021 report by the Arctic Council, and that warming is causing the ice to expand and contract in unpredictable ways.
Some scientists and research companies are now using artificial intelligence-powered tools to provide more accurate and timely predictions about which parts of the Arctic Ocean will be covered by ice and when. AI’s algorithms complement existing models that use physics to understand what is happening on the surface of the ocean, a dynamic zone where cold underwater currents meet with harsh winds to create floating ice rafts. This information is becoming increasingly valuable to members of Arctic tribes, commercial fishermen in places like Alaska, and global shipping companies who are interested in shortcuts through open water.
Leslie Canavera, CEO of Polarctic-based Lorton, Virginia, a scientific consulting firm that has developed AI-based forecasting models, says the uncertain pace of climate change means that existing sea ice patterns are becoming less accurate. This is because they are based on processes in the environment that are changing rapidly.
“We do not have an excellent understanding of climate change and what is happening in the country [Arctic] system, ”said Canavera, a member of the Yupik tribe who grew up in Alaska. “We have statistical modeling, but then you look at more than the average. Then you have artificial intelligence that can see the trends in the system and learn. “
Existing physics-based models capture hundreds of years of scientific records of ice conditions, current weather conditions, the speed and location of the polar jet stream, the amount of cloud cover, and ocean temperature. The models use this data to estimate future ice cover. But it takes large amounts of computing power to add the numbers, and a few hours or days to make a prediction using conventional programs.
While AI also requires complex data and very initial computing power, once the algorithm is trained in the right amount and type of data, it can detect climate models faster than physics-based models, according to Thomas Anderson, a British Antarctic data scientist. Survey, which developed an AI forecast for ice called IceNet. “AI methods can just work thousands of times faster, as we found in our model, IceNet,” says Anderson. “And they also learn automatically. AI is no smarter. It does not replace physics-based models. I think the future uses both sources of information. “
Anderson and colleagues published their new model for sea ice forecast in August in the magazine Natural communications. IceNet uses a form of AI called in-depth training (also used to automate credit card fraud detection, driving self-driving cars, and managing personal digital assistants) to train to provide a six-month forecast in every 25-kilometer square network in the region, based on simulations of the Arctic climate between 1850 and 2100 and actual observational data recorded from 1979 to 2011. After the model was trained and given current weather and ocean conditions, IceNet defeated a leading physics-based model. in making seasonal forecasts of the presence or absence of sea ice in each square of the network, especially for the summer season, when the ice passes through an annual retreat, according to nature study.
[ad_2]
Source link