We obtained the global results using remote sensing reanalysis dataset. But the snow-coverd areas should be removed (like the blank area in the picture below, sources: Purdy et al., 2016). Does anyone know where we could obtain the snow data as mask? Thanks a lot!