A new AI tool could allow scientists to more accurately forecast Arctic sea ice months into the future. IceNet is almost 95% accurate in predicting whether sea ice will be present two months ahead, researchers say. It’s one of a growing number of uses for AI in predicting climate change. “AI has significantly improved the efficiency of running complex climate models that historically have been computationally intensive,” Daniel Intolubbe-Chmil, an analyst at Harbor Research,, told Lifewire in an email interview.
No Ice, Ice, Baby
IceNet is working on the formidable challenge of making accurate Arctic sea ice forecasts for the season ahead. Researchers described how IceNet works in a recent paper published in the journal Nature Communications. “Near-surface air temperatures in the Arctic have increased at two to three times the rate of the global average, a phenomenon known as Arctic amplification, caused by several positive feedbacks,” the researchers wrote in the paper. “Rising temperatures have played a key role in reducing Arctic sea ice, with September sea ice extent now around half that of 1979 when satellite measurements of the Arctic began.” Sea ice is hard to forecast because of its complex relationship with the atmosphere above and the ocean below, according to the paper’s authors. Unlike conventional forecasting systems that attempt to model the laws of physics directly, the researchers designed IceNet based on a concept called deep learning. Through this approach, the model “learns” how sea ice changes from thousands of years of climate simulation data, along with decades of observational data, to predict the extent of Arctic sea ice months into the future. “The Arctic is a region on the frontline of climate change and has seen substantial warming over the last 40 years,” the paper’s lead author, Tom Andersson, a data scientist at the BAS AI Lab, said in a news release. “IceNet has the potential to fill an urgent gap in forecasting sea ice for Arctic sustainability efforts and runs thousands of times faster than traditional methods.”
AI Casts a Broad Net
Other AI simulators are keeping an eye on climate change as well. Researchers have leveraged the Deep Emulator Network Search technique, for example, to improve a simulation around the way soot and aerosols reflect and absorb sunlight. The research found the emulator was 2 billion times faster and more than 99.999% identical to their physical simulation. AI and weather analytics also can help combat climate change by reducing emissions in the supply chain, Renny Vandewege, a vice president at the weather forecasting company DTN, told Lifewire in an email interview. “For example, in shipping, weather-optimized routing can reduce emissions up to 4% and reduce fuel consumption up to 10%, and weather routing in the aviation industry can prevent unnecessary re-routing to avoid bad weather, or circling an airport waiting to land,” he said. Precise forecasting for road networks can reduce unnecessary treatment of winter roads, reducing the number of harmful chemicals, Vandenwege said. “Instead of treating an entire roadway, road maintenance crews can choose to treat selected locations along a road where there are cold-spot road sections, or they may decide whether treatment is necessary at all,” he added. Machine learning and AI models are increasingly being used to help understand emissions of CO2 and Methane, Marty Bell, the chief science officer at weather forecasting company WeatherFlow, told Lifewire in an email interview. “The models are also increasing our resilience to climate change by helping us modify our approach to energy production and usage,” Bell said. “While many of these AI applications operate at large scales on utility energy distribution systems, others operate at the household level where ML informs AI models embedded in everyday internet-of-things devices that more efficiently manage energy usage in the house.”
