The challenge of monitoring snow surface sublimation in winter could be resolved with structure-from-motion photogrammetry

Snow sublimation is one of the major water and energy balance processes for the snow water cycle, which induced a significant snow loss. The precise monitoring of snow sublimation is essential for quantifying the snow processes. However, the conventional measurement methods (e.g., pan observation with sporadic data or eddy covariance [EC]) are not feasible for long-term observations in harsh alpine environments). Here, two different one-camera time-lapse structure from motion (O-T-SfM) photogrammetry methods, were adopted to monitor daily snow surface sublimation at two different periods in the cold, harsh conditions at top of the August-one ice cap in the Qilian Mountains of China. The results revealed that the O-T-SfM-based surface sublimation rate estimates has a good agreement with two different methods such as pan observation and EC measurement in snowfall-free conditions and snow bedforms were abundant. The estimated sublimation rate varied from 0.05 to 2.19 mm/day at the study site. The total surface sublimation was about 16.3 mm snow water equivalent in 32 days which is a significant loss of moisture from glacier in a short period. Our results suggest that by apply appropriate estimation methods, remote O-T-SfM photogrammetry derived DEM sequence can be successfully used to monitor the daily snow surface sublimation in winter.