Drivers of net glacier balance at high altitude in the Himalayas

Séminaire de M. Litt (Univ. Utrecht), le Jeudi 6 Février 2020 à 11h, salle Lliboutry, Bât. Glaciologie

Titre : Drivers of net glacier balance at high altitude in the Himalayas

Séminaire de : Maxime Litt, Post-doc à Utrecht University, Pays Bas

Abstract :
Climate is changing, inducing shifts into Glaciers mass balance and dynamics, and impacting snow cover. We want to document and understand the climate behavior in the Himalayas, with a precise interpretation of the spatio-temporal patterns of glacier changes. An important need in the region is a reliable evaluation of future snow melt water availability, in order to correctly forecast the seasonal river runoff. At high-altitudes in the Himalayas, net snow accumulation and surface energy balance respond to specific processes which require a specific understanding. We present some aspects of these processes, for the case of the Khumbu and the Langtang valleys in Nepal where in-situ meteorological data is available. First, we compare the precipitation outputs from downscaled WRF runs, obtained with either dynamical, either linear methods. This helps us to quantify the effect of the steep orography, which induces a complex interaction between atmospheric layers and challenges our ability to forecast total snow deposition with large-scale precipitation modelling. The analysis points out the role of the rain-gauges or snow sensor positions in obtaining reliable basin scale estimates. Second, we use in-situ data from various in-situ meteorological stations. Analysis of snow-drift sensor’s outputs highlights the importance of post-deposition wind-driven snow transport. We show that the calculated melt barely represents the actual ablation. Third, a statistical study of daily meteorological station data shows that low air-density reduces heat exchanges between ice and air and favors sublimation. At daily scale, the total energy available for melt decorrelates from surface air temperature and the degree-day approach fails. We propose an alternative to this approach.