Photo Credit: WAM
Researchers at NYU Abu Dhabi have developed a new forecasting algorithm capable of predicting Arctic sea ice extent up to nine months in advance, potentially helping scientists better understand climate changes with global consequences.
The new method, called the Random Analogue Predictor (RAP), was developed by researchers at the Mubadala Arabian Centre for Climate and Environmental Sciences (ACCESS) at NYU Abu Dhabi.
RAP analyses historical records of Arctic sea ice and identifies past conditions that closely resemble the current situation. It then examines how those earlier conditions developed to produce a range of possible future forecasts.
The system also estimates the level of uncertainty associated with each prediction, giving researchers a clearer indication of how much confidence they can place in the forecast.
Arctic sea ice is an important component of the global climate system. Its bright surface reflects solar energy back into space, while darker ocean water absorbs more heat. Changes in sea ice levels can also affect atmospheric and ocean circulation patterns well beyond the Arctic.
Francesco Paparella, inaugural director of Mubadala ACCESS at NYU Abu Dhabi and senior author of the study, said forecasting Arctic sea ice months in advance remains challenging as the region continues to experience rapid environmental changes.
He said the new approach was intentionally designed to remain relatively simple while delivering results comparable with considerably more complex forecasting systems.
According to the researchers, RAP demonstrated forecasting performance similar to models participating in the Sea Ice Prediction Network (SIPN). For forecasts of September sea ice extent, its prediction error was comparable to that of 34 seasonal forecasting models.
Unlike physics-based climate models that simulate interactions between the atmosphere, ocean and sea ice, RAP relies solely on historical records of Arctic sea ice extent.
By identifying comparable conditions from previous years, the algorithm generates multiple possible forecasts. Differences between those forecasts are then used to estimate uncertainty.
Researchers said RAP could serve as a benchmark for evaluating both physics-based forecasting systems and emerging artificial intelligence models.
Its relatively simple and transparent design could provide scientists with a useful reference point when assessing whether more sophisticated forecasting techniques deliver meaningful improvements in predicting future Arctic sea ice conditions.
