| Summary: | Overview We provide static predictions of net snow accumulation over the Antarctic Ice Sheet derived using amachine learning approach.Here, we train random forest models to predict variability in net accumulation using atmospheric variables and topographic characteristics as predictors at 1 km resolution. Observations of net snow accumulation from both in situ and airborne radar data provide the input observable targets needed to train the random forest models. The data file includes all the predictors and predictands, an independent stake transect used for evaluation, four random forest gridded accumulation anomalies and their uncertainties, and a combined accumulation anomalies and its uncertainty. For a thorough description of how the predictions were generated generated see Medley et al. (2022).
|
|---|