Distributed Melt on a Debris-Covered Glacier: Field Observations and Melt Modeling on the Lirung Glacier in the Himalaya

Jakob F. Steiner,Walter W. Immerzeel,Philip D. A. Kraaijenbrink

doi:10.3389/feart.2021.678375

Jakob F. Steiner, Walter W. Immerzeel + Show 1 more

Open Access

https://doi.org/10.3389/feart.2021.678375

Copy DOI

Abstract

Debris-covered glaciers, especially in high-mountain Asia, have received increased attention in recent years. So far, few field-based observations of distributed mass loss exist and both the properties of the debris layer as well as the atmospheric drivers of melt below debris remain poorly understood. Using multi-year observations of on-glacier atmospheric data, debris properties and spatial surface elevation changes from repeat flights with an unmanned aerial vehicle (UAV), we quantify the necessary variables to compute melt for the Lirung Glacier in the Himalaya. By applying an energy balance model we reproduce observed mass loss during one monsoon season in 2013. We show that melt is especially sensitive to thermal conductivity and thickness of debris. Our observations show that previously used values in literature for the thermal conductivity through debris are valid but variability in space on a single glacier remains high. We also present a simple melt model, which is calibrated based on the results of energy balance model, that is only dependent on air temperature and debris thickness and is therefore applicable for larger scale studies. This simple melt model reproduces melt under thin debris (&lt;0.5 m) well at an hourly resolution, but fails to represent melt under thicker debris accurately at this high temporal resolution. On the glacier scale and using only off-glacier forcing data we however are able to reproduce the total melt volume of a debris-covered tongue. This is a promising result for catchment scale studies, where quantifying melt from debris covered glaciers remains a challenge.

Highlights

In this study we combine modeling with a unique number of observed datasets from the debris-covered Lirung Glacier, to quantify mass loss in time and space using an energy balance model as well as two index models
The two overarching goals were to evaluate the performance of an energy balance model to reproduce melt at an hourly resolution over a debris-covered tongue, and test the applicability of a computationally fast, but simpler index model to be applied at larger scale
Comparison to in-situ thickness measurements show this to be a viable approach for the glacier scale, with a mean absolute error (MAE) of 0.22 m and a mean bias error (MBE) of −0.005 m

Summary

Introduction

Debris-covered glaciers are common in a number of glaciated mountain ranges, including highmountain Asia [HMA, (Scherler et al, 2011)], the European Alps (Brock et al, 2010), the Caucasus (Lambrecht et al, 2011), the Andes of Chile (Janke et al, 2015) and Peru (Wigmore and Mark 2017), and the Russian (Barr et al, 2018), North American (Herreid and Pellicciotti 2018) and Scandinavian Arctic (Midgley et al, 2018). Studies in the Alps and HMA have shown that energy balance models perform well to estimate melt at the point scale These models are very sensitive to thickness, thermal conductivity and surface roughness of the debris layer, all of which are difficult to measure (Nicholson and Benn 2006; Reid and Brock 2010; Rounce et al, 2015). A number of studies have observed and modeled melt of supraglacial ice cliffs (Sakai et al, 2002; Steiner et al, 2015; Brun et al, 2016; Buri et al, 2016a) and ponds (Sakai et al, 2000; Miles and Arnold, 2016; Watson et al, 2016), which are common surface features on debris-covered tongues, especially in the Himalaya. Around these features melt intensifies considerably, but they cover only a relatively small area of the total tongue (Steiner et al, 2019)

Objectives

Methods

Results

Conclusion