What Can Thermal Imagery Tell Us About Glacier Melt Below Rock Debris?

Abstract

Rock debris on the surface of a glacier can dramatically reduce the local melt rate, where the primary factor governing melt reduction is debris layer thickness. Relating surface temperature to debris thickness is a recurring approach in the literature, yet demonstrations of reproducibility have been limited. Here, I present the results of a field experiment conducted on the Canwell Glacier, Alaska, United States to constrain how thermal data can be used in glaciology. These datasets include, 1) a measured sub-daily “Østrem curve” time-series; 2) a time-series of high resolution thermal images capturing several segments of different debris thicknesses including the measurements from 1); 3) a thermal profile through a 38 cm debris cover; and 4) two Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) satellite thermal images acquired within 2 and 3 min of a field-based thermal camera image. I show that, while clear sky conditions are when space-borne thermal sensors can image a glacier, this is an unfavorable time, limiting the likelihood that different thicknesses of debris will have a unique thermal signature. I then propose an empirical approach to estimate debris thickness and compare it to two recently published methods. I demonstrate that instantaneous calibration is essential in the previously published methods, where model parameters calibrated only 1 h prior to a repeat thermal image return diminished debris thickness estimates, while the method proposed here remains robust through time and does not appear to require re-calibration. I then propose a method that uses a time-series of surface temperature at one location and debris thickness to estimate bare-ice and sub-debris melt. Results show comparable cumulative melt estimates to a recently published method that requires an explicit/external estimate of bare ice melt. Finally, I show that sub-pixel corrections to ASTER thermal imagery can enable a close resemblance to high resolution, field-based thermal imagery. These results offer a deeper insight into what thermal data can and cannot tell us about surface debris properties and glacier melt.

Using near–surface temperature data to vicariously calibrate high-resolution thermal infrared imagery and estimate physical surface properties

“…More recently, thermal imaging studies have expended temporal coverage. Herreid [14] used three-hour increments over multiple days to determine glacier melt under rock debris. High temporal resolution, such as this study, typically focus on short-term phenomena, such as cooling lava flows [33] .…”

“…The use of thermal cameras to analyze landscapes has expanded over the last decade (e.g., [ 6 , 14 , 15 , 21 , [32] , [33] , [34] , 39 , 40 ]). However, depending on the type and age of the thermal cameras, the output image may be self-calibrated in radiance, corrected for atmospheric absorption and emission, and each pixel converted to surface temperature values.…”

Using ground-based thermal imagery to estimate debris thickness over glacial ice: fieldwork considerations to improve the effectiveness

“…At this resolution, local slope and aspect variations, which play an important role in controlling the surface energy balance and therefore the surface temperature, are not captured. Due to the large pixel area of these thermal images, debris, ice cliffs and supraglacial ponds can all appear in a single pixel, resulting in an underestimation of the debris thickness (Rounce and others, 2018; Herreid, 2021). Nicholson and others (2018) showed that ignoring the small-scale variability in debris thickness and using spatially averaged values can have a strong impact on the melt predicted by glacier models, underestimating the ablation rate by 11–30%.…”

“…One option to capture surface temperature at a high spatial resolution in glacierized terrain is to use near-surface remote sensing, such as from a plane or uncrewed aerial vehicles (UAVs, Kraaijenbrink and others, 2018) or ground-based oblique imagery (Hopkinson and others, 2010; Aubry-Wake and others, 2015; Herreid, 2021; Tarca and Guglielmin, 2022). These instruments offer the possibility to measure surface temperatures from thermal infrared radiometry (TIR) both at a high spatial resolution and with flexible timing to account for different times of the day or variable weather conditions.…”

The Infierno Glacier (Pyrenees, Aragon, Spain): evolution 2016-2022

“…This phenomenon might be related to the progressive melting of permafrost at these altitudes and to increasing periglaciar dynamics (Figure 4). It is acknowledged that the upper covers of washout located at some centimeters isolate the ice and reduce the melting process [40][41][42][43][44]. In spite of it, these glaciers are becoming thinner, their velocity is modified, and it has been observed a compression of the ice through the weight of the melt [45].…”