Using climate reanalysis data in conjunction with multi-temporal satellite thermal imagery to derive supraglacial debris thickness changes from energy-balance modelling

Rebecca L. Stewart,John Woodward,Francesca Pellicciotti,Darrel Swift,Ann Rowan,Benjamin Brock,Matthew Westoby

doi:10.1017/jog.2020.111

Abstract

AbstractSurface energy-balance models are commonly used in conjunction with satellite thermal imagery to estimate supraglacial debris thickness. Removing the need for local meteorological data in the debris thickness estimation workflow could improve the versatility and spatiotemporal application of debris thickness estimation. We evaluate the use of regional reanalysis data to derive debris thickness for two mountain glaciers using a surface energy-balance model. Results forced using ERA-5 agree with AWS-derived estimates to within 0.01 ± 0.05 m for Miage Glacier, Italy, and 0.01 ± 0.02 m for Khumbu Glacier, Nepal. ERA-5 data were then used to estimate spatiotemporal changes in debris thickness over a ~20-year period for Miage Glacier, Khumbu Glacier and Haut Glacier d'Arolla, Switzerland. We observe significant increases in debris thickness at the terminus for Haut Glacier d'Arolla and at the margins of the expanding debris cover at all glaciers. While simulated debris thickness was underestimated compared to point measurements in areas of thick debris, our approach can reconstruct glacier-scale debris thickness distribution and its temporal evolution over multiple decades. We find significant changes in debris thickness over areas of thin debris, areas susceptible to high ablation rates, where current knowledge of debris evolution is limited.

Highlights

Debris-covered glaciers are found in most glacierised regions (Reid and others, 2012; Kirkbride and Deline, 2013; Anderson and Anderson, 2018; Scherler and others, 2018)
As the model is sensitive to changes in wind speed, until a reanalysis product emerges which is able to reliably predict surface level wind speed, we suggest that a fixed value should be derived using the methodology presented above, or in the absence of available Automatic Weather Station (AWS) data, a fixed value of 2.19 m s−1 should be used for alpine settings such as Miage Glacier and Haut Glacier d’Arolla, and a value of 1.41 m s−1 should be used for Himalayan settings
Using ERA-5 reanalysis data as the meteorological forcing, we present a ∼20-year time series of debris thickness estimates at each study glacier, and an assessment of debris thickness changes that have occurred during the study period

Summary

Introduction

Debris-covered glaciers are found in most glacierised regions (Reid and others, 2012; Kirkbride and Deline, 2013; Anderson and Anderson, 2018; Scherler and others, 2018). As debris thickness increases from zero on a clean-ice surface, the sub-debris melt rate will increase until the critical thickness is reached, representing the maximum ablation rate. The maximum ablation rate is typically higher than for a climatologically equivalent clean-ice surface (Østrem, 1959; Mattson and others, (1993); Evatt and others, 2015). Kirkbride and Deline, 2013; Scherler and others, 2018; Tielidze and others, 2020), it is imperative to improve our understanding of, and ability to model the effects of surface debris on key processes affecting debriscovered glaciers, most notably ablation As the areal extent of debris-covered ice increases globally (e.g. Kirkbride and Deline, 2013; Scherler and others, 2018; Tielidze and others, 2020), it is imperative to improve our understanding of, and ability to model the effects of surface debris on key processes affecting debriscovered glaciers, most notably ablation

Methods

Results

Discussion

Conclusion