From the Ground Up: Integrating Field Measurements With Remote Sensing for Improved Glacier Evolution Modeling

Abstract

Glaciers around the world are losing mass at unprecedented rates, with profound consequences for sea level rise, freshwater resources, natural hazards, and downstream human communities. Among all glaciated regions, Alaska stands out as a critical hotspot of glacier change: its extensive glacierized area contributes disproportionately to global sea level rise and observed mass loss rates are accelerating faster than in any other major glaciated region excluding the ice sheets. Still, glaciers in Alaska remain relatively under-studied and substantial uncertainties remain in existing large-scale observational datasets, thus limiting our process-based understanding of ongoing changes and ability to model future changes. Addressing these uncertainties requires a combination of “ground truth” field observations, enhanced large-scale observational datasets, and glacier evolution model developments that can integrate disparate observational datasets.

I develop novel approaches to measure daily mass balance and glacier dynamics of Gulkana Glacier, Alaska; this includes the first application of GNSS reflectometry for high-resolution monitoring of mountain glaciers. We find subseasonal patterns in melt and dynamics across the glacier, the latter of which is driving by subglacial hydrology. We compare this data with existing data products for Alaska, ultimately showing that existing large-scale products are unable to capture the spatially- and temporally-varying mass balance and dynamics of Gulkana Glacier. This challenge exists for glaciers across the broader Alaska region, where similar improvements to existing observations are necessary to understand complex glacier processes at sufficient resolutions. Overcoming these challenges lies within leveraging field data as ground-truth for spatially-expansive remote sensing or as calibration within high-order physics-based models.

While the fieldwork highlighted the need for better remote sensing products for scaling, I also sought to determine how historical datasets could be used to inform our understanding of ongoing and future changes. In the second chapter of this thesis, I expand the temporal extent of observations using historical aerial photography on Kennicott and Root glaciers, Alaska. Such historical observations provide rare constraints on glacier behavior prior to the satellite era and inform comparisons with contemporary changes. We observe near-equilibrium conditions prior to 1957, followed by accelerating mass loss since. This observed long-term record of mass loss is defined by patterns of terminus slowdown and eventual stagnation that coincides with glacier thinning and debris-cover expansion. Ultimately, we show that historical records can be valuable assets towards improving glacier projections. We find that Kennicott and Root glaciers–two hallmarks of the Wrangell-St. Elias National Park and Preserve–are projected to lose 38% to 63% of their mass during the 21st century, depending on the future climate.

After investigating the importance of observations at various temporal scales for a few glaciers, I aimed to gain a more holistic view of changes across the Alaska region. To that end, I process spatially extensive remote sensing data from mid-2016 through 2024 to characterize glacier response to changes in climate across Alaska. Specifically, we derive transient snowlines and melt extents–two insightful proxies of glacier mass balance–for all glaciers in Alaska with an area of at least 2 sq. km. We find that +1˚C of summer warming results in up to 3 additional weeks of glacier melt, and an extreme heat wave in 2019 caused snowlines retreat that exposed 28% more underlying ice than typical years.

Finally, I set out to understand the impacts and implications of these large-scale observations on the future of Alaskan glaciers. As such, I calibrate a large-scale glacier evolution model with melt extent and snowline observations to improve projections of glacier mass loss for all glaciers in Alaska to 2100. We find that the model calibrated with transient snowlines and melt extents shows strong agreement (r-squared=0.82) with snowline observations and glaciological seasonal mass balance measurements. Depending on the future climate scenario, we project Alaska to lose 49 ± 17% to 73 ± 26% of its mass by 2100, relative to 2015. We show that these changes are accompanied by rising equilibrium-line altitudes, increased surface melt, and a shift from snow to rain precipitation in the late summer months.

Collectively, this dissertation evaluates the complementary value of glacier observations acquired across a wide range of spatial and temporal scales, and examines how each contributes to our understanding of glacier change. By integrating novel field measurements, historical reconstructions, and contemporary remote sensing within a unified modeling framework, this work highlights the distinct insights provided by different data types and resolutions, as well as their limitations. The results demonstrate how strategically combining these observations improves process understanding, constrains key sources of uncertainty, and strengthens projections of future glacier change. In doing so, this dissertation advances a more comprehensive and data-informed perspective on Alaskan glacier response to a warming climate.