Skip to main content

Advancing Snow Modeling

The Snow Water Equivalent Machine Learning(SWEML) incorporates ground-based snow measuring sites, remotely-sensed snow cover information, and a Artificial Neural Network to provide point estimations of Snow Water Equivalent. The network was trained on historical data data from NASA’s ASO missions, divided into regions, and then a LightGradientBoost Model was used to preform recursive feature elimination to produce an efficient feature selection and region-specific model. The class contains the required functions for downloading data, pre-processing, running inference, and for producing visualizations.