Provenance and transport of supraglacial debris revealed by variations in debris geochemistry on Khumbu Glacier, Nepal Himalaya

The origin of supraglacial debris covers is often conceptualised as the formation of a surface lag by melt‐out of englacial debris from slow‐moving ice, where complexity arises from feedback between debris thickness and sub‐debris ice melt. Here, we examine the origin of a debris cover from the perspective of debris provenance and changing tributary supply in a high‐elevation compound valley glacier. Geochemical analysis of 11 major elements in 21 debris samples from six tributaries of Khumbu Glacier (Nepal) shows unambiguous statistical differentiation of debris sources reflecting lithological differences between tributary catchments. Twenty‐four samples from transects across the ablation area are partitioned according to their source areas using the FR2000 sediment unmixing model. We estimate the age of ice at each transect using a higher order ice flow model. The results show greater proportions of debris from lateral tributaries in downglacier locations that have experienced longer flowline histories. More recently, ice from the Main Himalayan Divide (Western Cwm) has become relatively more important. This suggests a change in the state of the lower glacier's structure depending on the relative ice discharges of lateral and divide sources. Ice flux from lower elevation tributaries was more important probably prior to a weakening of the Indian Summer Monsoon at around 1420 CE. The lower elevation tributaries lie within the range of late Holocene equilibrium line altitude variation and therefore respond most sensitively to climatic drivers of the glacier's flow structure. Negative glacier mass balance since around 1900 CE caused tributary glaciers to detach and high‐elevation catchments to re‐establish as the dominant ice source to Khumbu Glacier. Geochemical patterns in the debris cover of a compound Himalayan glacier are matched to source catchments to understand changes over recent centuries in tributary glacier fluxes using sediment unmixing and glacier flow modelling.