Timeline gaps of reflectiveness of Arctic sea ice were bridged by using machine learning
Great challenges for the optical satellite measurements of sea ice in the Arctic include observations that were missing owing to a lack of sunlight, as well as cloudiness. A machine learning method known as gradient boosting allows the timeline of the reflectiveness, or albedo, of Arctic sea ice to be augmented by nearly one third.
In the research, the augmentation is based on the monthly averages of the albedo. Also used in cloudy and dark conditions were daily radiant temperature levels in the microwave area, and sea ice maps based on them.
Further information:
Emmihenna Jääskeläinen, Finnish Meteorological Institute, emmihenna.jaaskelainen@fmi.fi
Jääskeläinen, E., Manninen, T., Hakkarainen, J. and Tamminen, J.: Filling gaps of black-sky surface albedo of the Arctic sea ice using gradient boosting and brightness temperature data, International Journal of Applied Earth Observation and Geoinformation, 107,102701 (2022). doi:/10.1016/j.jag.2022.102701.
Scientific article is available on International Journal of Applied Earth Observation and Geoinformation.