Between Rain and Snow, Machine Learning Finds 9 Precipitation Types

Using nearly a decade of high-speed camera footage, University of Michigan researchers led a NASA-funded study that identified nine distinct types of precipitation, aided by machine learning. By analyzing 1.5 million minutes of data from sites across North America and Europe, they developed a model that accurately captures transitions between rain, snow, and mixed precipitation. Read the full story to learn how this research could improve weather forecasting and winter road safety.