Identifying and mapping very small (&lt;0.5 km2) mountain glaciers on coarse to high-resolution imagery

J. R. Leigh,D. J. A. Evans,L. M. Andreassen,I. S. Evans,R. J. Carr,C. R. Stokes

doi:10.1017/jog.2019.50

Abstract

AbstractSmall mountain glaciers are an important part of the cryosphere and tend to respond rapidly to climate warming. Historically, mapping very small glaciers (generally considered to be <0.5 km2) using satellite imagery has often been subjective due to the difficulty in differentiating them from perennial snowpatches. For this reason, most scientists implement minimum size-thresholds (typically 0.01–0.05 km2). Here, we compare the ability of different remote-sensing approaches to identify and map very small glaciers on imagery of varying spatial resolutions (30–0.25 m) and investigate how operator subjectivity influences the results. Based on this analysis, we support the use of a minimum size-threshold of 0.01 km2for imagery with coarse to medium spatial resolution (30–10 m). However, when mapping on high-resolution imagery (<1 m) with minimal seasonal snow cover, glaciers <0.05 km2and even <0.01 km2are readily identifiable and using a minimum threshold may be inappropriate. For these cases, we develop a set of criteria to enable the identification of very small glaciers and classify them ascertain,probableorpossible.This should facilitate a more consistent approach to identifying and mapping very small glaciers on high-resolution imagery, helping to produce more comprehensive and accurate glacier inventories.

Highlights

Numerous studies have revealed unprecedented global glacier recession during the late 20th and early 21st century, which has been linked to anthropogenically-induced climate change (e.g. Haeberli and others, 2007; Marzeion and others, 2014; Zemp and others, 2015)
Other recent glacier inventories such as the Inventory of Norwegian Glaciers (Andreassen and others, 2012b) have utilised orthophotographs as a means of validating their glacier maps which were compiled from satellite imagery
The results of the different mapping techniques applied to different imagery are displayed in Table 3, which includes summary statistics of the glaciers mapped in the Norwegian Glaciers (NGI) (Andreassen and others, 2012b)

Summary

Introduction

Numerous studies have revealed unprecedented global glacier recession during the late 20th and early 21st century, which has been linked to anthropogenically-induced climate change (e.g. Haeberli and others, 2007; Marzeion and others, 2014; Zemp and others, 2015). There has been a significant mass loss from the large polar ice sheets (Shepherd and others, 2012), the combined melt from mountain glaciers and ice caps between 2003 and 2009 accounted for 29 ± 13% of observed sea level rise, approximately equal to the loss from both the Greenland and Antarctic ice sheets (IPCC, 2013). It has been proposed that the combined melt from ‘uncharted glaciers’ (i.e. glaciers that are not currently included in global glacier inventories) may account for as much as 42.7 mm (31% of a total 137.1 mm sea level equivalent from glaciers globally, excluding the Greenland and Antarctic ice sheets) of sea level rise between 1901 and 2015 (Parkes and Marzeion, 2018). Other recent glacier inventories such as the Inventory of Norwegian Glaciers (Andreassen and others, 2012b) have utilised orthophotographs as a means of validating their glacier maps which were compiled from satellite imagery

Objectives

Methods

Results

Discussion

Conclusion