Code and Data for "A machine learning approach for estimating snow depth across the European Alps from Sentinel-1 imagery"

Here we share the data and code for “A machine learning approach for estimating snow depth across the European Alps from Sentinel-1 imagery”

Corresponding author: Devon Dunmire [e-mail]

‘model_training’ - contains script to train the ML model, and training data sets from (1) in-situ snow measurement sites (training_data.p) and (2) photogrammetry snow depth maps (map_training_data.p)

‘Cross_val_predictions’ contains model predictions for our cross-validation of all the in-situ snow measurement sites

‘run_model’ contains the trained model (final_model_xg.pkl) and scripts to retrieve snow depth with our ML model.

‘SD_*’ zip folders contains daily ML snow depth output over the European Alps for each snow year from Sept. 1 2015 - Apr. 30 2023. Data from multiple orbits is averaged.

Naming convention: ‘S1_ml_SD_{yyyymmdd}_.nc’