A new methodology for measuring ice cliff backwasting rates on debris-covered glaciers using high-resolution unmanned aerial vehicle imagery

Abstract

Ice cliffs potentially contribute considerably to the glacier mass balance, as melt enhancement on ice cliffs was observed by multiple recent studies. However, quantification of ice cliff melt is still in its infancy. Distributed models are still very computationally intensive and no methodology exists that allows for the direct measurement of ice cliff backwasting. This study therefore developed a new methodology to directly measure ice cliff backwasting on high-resolution UAV imagery with the Multiscale Model to Model Cloud Comparison (M3C2) algorithm. Two sets of experiments showed that a normal scale of 20 m in combination with a horizontal constraint of the normals in average backwasting direction resulted in accurate cliff backwasting values. The technique allowed cliff-to-cliff measurements with a RMSE of only 0.4 m. Backwasting patterns were generated for five cliffs on the Nepalese Langtang glacier for the period May 2014 - October 2015, which revealed an average backwasting rate of 10.5 m a-1. The rate is 13 times higher than the average ablation of 0.8 m a-1 on this part of the glacier tongue and could be an important explanation for the observed debris-cover anomaly. The backwasting rate varied considerably among different parts of the cliffs and between different cliffs, which could be well explained by the influence of aspect and supraglacial ponds. Supraglacial ponds caused a consistent positive melt gradient from cliff top to base, indicating a large role of thermal erosion in the ablation and development of ice cliffs. The M3C2 algorithm was applied to an additional cliff on Lirung glacier, which showed similar backwasting values as found by previous studies on that cliff. The methodology developed in this study is faster and probably less complex than other studies measuring or modelling ice cliff backwasting. The temporal and spatial scalability of the methodology will improve accurate assessment of a glacier’s mass balance: the melt contribution of ice cliffs is finally revealed through detailed backwasting patterns. Automatization of the different steps in this research is recommended to facilitate this process.