Is Artificial Intelligence The Future of Avalanche Forecasting? - SnowBrains

Considered by some observers as more of an art than a science, avalanche forecasting involves a unique combination of human observation, analysis, and interpretation. Even the best forecasters only reach about 75% accuracy in their predictions. After all, they are human too and susceptible to bias and imperfection just like the rest of us. One of the biggest challenges in creating an accurate forecast lies in the fact that avalanche danger cannot be precisely measured and is therefore a matter of expert assessment (opinion). That was until last season when the Swiss Institute for Snow and Avalanche Research Group (SLF), a part of the National Swiss Federal Institute for Forest, Snow, and Landscape Research successfully tested a first-of-its-kind, artificial intelligence computer program to assist in its avalanche forecasts.