Assessing the impact of terrestrial photography-derived roughness lengths on evaposublimation from seasonal snow in the Mediterranean mountains of Sierra Nevada (Spain)

1. Study Region The Sierra Nevada mountain range in Spain. 2. Study focus We estimate time series of snow aerodynamic roughness length from daily terrestrial photographs, leveraging textural features to analyze its influence on evaposublimation estimates in our regional snow model under an isotropic fetch assumption. Over a long-term study (2010–2021), we compare our derived roughness length time series with a constant roughness length, examining their implications for model performance and evaposublimation trends. 3. New hydrological insights for the region Our comparison reveals that the estimated contribution of evaposublimation to total annual snow ablation using roughness length time series is significantly higher (up to $(48.8\pm 3.6)\text{\%}$) than previously reported values ($30\text{–}35\text{\%}$) and improve performance of our regional snow model upon using roughness length time series. In our study period spanning twelve years, cumulative evaposublimation decline ($\text{-7.23}\text{mm/yr}$), likely linked to decreasing snow cover duration ($\text{-4.57}\text{d/yr}$) and decreasing annual solid precipitation ($\text{-24.81}\text{mm/yr}$), though no trend in evaposublimation rates is observed. The Sierra Nevada’s shallow snowpacks and multiple accumulation and ablation periods create mixed surface types with patchy snow, rocks, and vegetation influencing roughness lengths. Consequently, we identify a relationship between roughness length and snow cover fraction, with the potential to facilitate evaposublimation rate estimates on a spatio-temporal scale for this region in the future.