Research
Research statement
I utilize remote sensing, numerical modeling, and field data to address scientific questions within glaciology and earth surface processes. I am eager to involve undergraduates in cutting-edge research in collaboration with leading scientists from around the world. I have successfully built a research program propelled by substantive student theses that build students' quantitative, geospatial, and critical thinking skills. If you are interested in such research, please contact me at armstrongwh@appstate.edu
I consider myself a broad and engageabe scientist and have investigated a range of topics throughout my academic career including: mechanics driving glacier motion, downstream impacts of glacier change, scripted geospatial analysis, effects of climate change on flooding in the northeastern US, and rates of geomorphic change on a low-gradient alluvial river.
Part of what makes science fun for me is interaction with others, and I am proud of strong collaborations I have built with researchers at the University of Alaska Fairbanks, ETH Zurich, Colorado State University, Dartmouth University, University of Colorado at Boulder, and elsewhere. These collaborations help me hold research students to a higher standard and also provide a network of prospective graduate advisors.
Current research
Centennial-scale evolution of glacier basal motion during climate warming
Glacier basal motion comprises sliding at the ice-bedrock interface or deformation within till underlying the glacier (see figure). Glacier basal motion is known to depend sensitively on glacier hydrology, but the exact mechanisms remain poorly understood. Glacier basal motion is responsible for the majority of ice motion on many glaciers, and is particularly important for ice flux on fast-moving tidewater glaciers and ice sheet outlet glaciers. Because basal motion is dependent on glacier hydrology, it can generate large changes in glacier speed over short timescales and allows glaciers to respond rapidly to environmental change. Further, basal motion is the means by which glaciers erode their beds, making this process critical for understanding how alpine and polar landscapes evolve over geologic time.
Recently, several studies have shown that glaciers move more slowly in hot years because they develop large rivers underneath the glacier that are able to effectively get rid of lubricating water; these results oppose earlier studies that suggested glaciers will move faster in a warmer climate. If glaciers slow down in a warming world, this could result in slower ice loss and sea level rise. However, it is difficult to make long-term predictions of glacier behavior from existing, relatively short datasets, because we know the way a glacier responds to meltwater reaching the bed can change over time.
Schematic depicting how an initially straight borehole is translated and deformed by glacier motion.
Our research, funded by the NSF Office of Polar Programs, is developing a unique long-term dataset to investigate how the speed and pattern of water-mediated glacier motion changes as a glacier’s shape and plumbing system adjust to a changing climate. Our work will allow better prediction of future glacier change and enhance understanding of how glaciers shape landscapes over geologic time. We are partnering with a documentary filmmaker and Jasper National Park (Alberta, Canada), which hosts our field site, to develop educational materials that will show park visitors the importance of glaciers and scientific research to human society.
Preliminary work utilizing limited field data combined with a wealth of historical and remotely-sensed data suggest that Athabasca Glacier's slowing over the last ~60 years is dominantly driven by declining basal motion due to the geomtry and hydrology-induced changes in the "stickiness" of the glacier bed (Armstrong et al., 2022).
On Athabasca Glacier with undergraduate researchers Hannah ('20, left) and Anton ('20, right). In March 2019 we installed a weather station that will provide data to model meltwater and rain inputs to Athabasca Glacier's subglacial "plumbing system". Photo by Greg Horne (Parks Canada).
This project is in collaboration with Martin Truffer and David Polashenski (U Alaska Fairbanks), who used a radar to measure Athabasca Glacier ice thickness in March 2019. These data will be used to site boreholes to the glacier bed and are needed for modeling glacier flow. Photo by Greg Horne.
a) Cross-valley plot of glacier velocity (gray contours) and instrument location (red dots). b) Down-borehole instrumentation will document the distribution of glacier basal motion, internal deformation, and subglacial water pressure.
Evolution of Alaska's ice-marginal lakes
Glaciers are prolific sediment producers and their connectivity with the proglacial environment affects downstream geomorphic processes, aquatic habitat, and nutrient transport. As glaciers retreat, they often leave behind large proglacial lakes that trap sediment, alter downstream hydrology, and may affect upstream glacier mass balance. In the face of climate warming, I ask: How are Alaska's ice-marginal lakes changing? Can we identify environmental predictors for observed lake change?
Student research
I have worked with 6 undergraduate students on a range of satellite-based ice-marginal lake projects to produce 2 peer-reviewed publications and >10 American Geophysical Union Fall Meeting presentations (as of 2022). Two examples are shown below.
Hannah Field ('20) explains two distinct trends we observe in satellite records of proglacial lake area change over the past 30 years.
Anton Hengst ('20) describes his approach to automate proglacial lake area delineation from satellite imagery.
We used repeat satellite imagery to investigate area changes on lakes in direct contact with glaciers across the Gulf of Alaska over 1984 - 2018 (Field et al.,2021; Rick et al., 2022). In general, we found that lakes downstream from glaciers are growing while lakes dammed by glaciers are shrinking. We considered a range of environmental variables (i.e., climate, glaciers, topography) to find which best predicted lake area change. Our research suggests that the shape of the landscape surrounding a lake is more important for driving the lake's growth rate than climatic or glaciologic factors.
The figure on the left shows area change on glacier-connected lakes over the study period, where green circles show growing lakes and red circles indicate shrinking lakes.
Related undergraduate-led research highlighted the necessity of utilizing high temporal resolution imagery for accurate characterization of ice-marginal lake area change (Hengst et al., 2022).
Upstream & downstream impacts of glacier retreat
Proglacial lakes (i.e., lakes that are in direct contact with a glacier terminus) are a fascinating and important research subject because they integrate many physical systems (atmosphere, cryosphere, hydrosphere) and are rapidly changing. My students and I investigate how the formation and growth of proglacial lakes alters the upstream glacier and downstream hydrologic and geomorphic systems. We use a wide range of satellite data and geospatial products to investigate changes in suspended sediment & channel morphology downstream from retreating glaciers. We have employed similar datasets to characterize the role of proglacial lakes in exacerbating glacier change. These studies are important for understanding and predicting how the world's glaciers will continue to change over the 21st century and how their change affects downstream habitat.
Previous research
Quantifying the spatial distribution of glacier basal motion on a regional scale
The mechanics of glacier basal motion are poorly understood partly because of lacking systematic observations across the entire length of a range of glaciers. Such observations will allow characterization of first-order controls on glacier basal motion. I generated a novel data processing routine to ingest the entire Landsat 8 archive to develop seasonal velocity profiles across 19 land-terminating Alaska glaciers. I found the summer speedup, which reflects enhanced summer basal motion, varies only by a factor of 2 across a range of glaciers, is independent of annual average speed, and is relatively uniform across the lower reach of the glaciers.
This work suggests that enhanced summer basal motion is similar across a wide range of glaciers, a self-regulation mechanism in the hydro-sliding process and may provide a mechanism for the formation of icefalls.
To learn more, read Armstrong et al. [2017] in Geophysical Research Letters.
Glacier velocity across the Wrangell-St Elias ranges of south-central Alaska and the western Yukon. Hot (cool) colors indicate fast (slow) glacier motion. This map was generated by calculating the median velocity at each pixel from over 30 satellite image pairs from 2013-2015. Figure from Armstrong et al. [2017]
Observing and modeling seasonal glacier speed change
We utilized feature tracking on high resolution (0.5 m pixel) WorldView satellite imagery to document the seasonal cycle of ice surface velocity on Kennicott Glacier, Alaska (see map below for location). These data provided spatial context for point observations of glacier speed made using on-glacier GPS (read more below). The processes we are interested in, however, are concealed from surface observation by over 500 m of viscously deforming ice that distorts the pattern of glacier basal motion. To investigate these hidden processes, we employed a numerical flow model to determine the possible distributions of glacier basal motion that could produce our surface observations.
This work suggests the possibility of simple parameterizations in glacier flow models and provides a context for interpreting field and remote sensing studies.
To learn more, read Armstrong et al. [2016] in Journal of Glaciology.
Observing and modeling seasonal glacier speed change. a) Observed (pink and gray bands) and modeled (solid lines) glacier surface speed produced by imposed basal velocity fields (dashed lines). b) Modeled cross-sectional profile of glacier velocity. Figure from Armstrong et al. [2016]
Map depicting instrument deployment on Kennicott Glacier.
Investigating the link between glacier hydrology and glacier motion
I deployed and maintained equipment to monitor daily-to-seasonal changes in the speed and hydrology of Kennicott Glacier, Alaska. The survey grade GPS units move with the glacier and are capable of determining their position on the surface of the earth to within 4 cm (1.6 in). We also measure ice surface melt and the water height of ice-dammed lakes, which reflects subglacial water pressure. We combine these observations to probe how subglacial water affects glacier basal motion.
To learn more, read Armstrong & Anderson [2020] in Journal of Glaciology.
A global positioning system (GPS) monument recording daily-to-seasonal speed changes related to glacier hydrology on Kennicott Glacier, Alaska.
Hydroclimatic flood trends
A warming global climate will increase the vigor of the hydrologic cycle. Several studies have documented increasing precipitation frequency and intensity, yet trends in streamflow are more ambiguous on a national scale.
With collaborators at the National Oceanic and Atmospheric Administration and Boston College, I conducted a series of studies documenting increasing trends in flood frequency (i.e., how often do floods occur) and flood magnitude (i.e., how large are floods when they happen) in the northeastern United States [Armstrong et al., 2012; 2014]. These trends, observed in watersheds with minimal direct human impact, are related to global climate change as well as low frequency variability in atmospheric circulation patterns.
Our findings are important for river restoration and aquatic habitat because changes in flood characteristics impact engineered structures and alter sediment mobility, which affects both ecosystem function and channel stability.
To learn more, read Armstrong et al. [2012] in the Journal of the American Water Resources Association and Armstrong et al. [2014] in Hydrological Sciences Journal.
Millennial river migration and sediment accumulation rates
Over geologic time, rivers wander across the landscape, resurfacing the land's surface and depositing nutrients required for wildlife and agriculture as they move. As a GeoCorps Guest Scientists, I collected and processed soil samples from ancient river channels to document how the Congaree River (South Carolina) has moved and accumulated sediment since the time the Greeks were recording Homer's Odessey.
Trends in the annual maximum flow (AMSQ) in the northeastern US. Blue (red) triangles indicate increasing (decreasing) trends, where the size of the triangle indicates the magnitude of the trend. Black circles indicate near-zero change. We found that floods are occurring more frequently in the Northeast, and, when they do occur, they are larger on average. Figure from Armstrong et al. [2014]
Surveying dam-induced sedimentation to generate a bsaeline dataset before dam removal on the Souhegan River, New Hampshire.
Research tools
Remote sensing
Remote sensing is the science of collecting data without physically coming into contact with your study subject. Remote sensing scientists investigate data collected by satellites, planes, and drones to study diverse topics such as earth surface processes, ecological health, humanitarian crises, ocean circulation, and more. I am broadly interested in developing creative new ways to use remotely-sensed data to better understand how our world works, with a particular focus on glacier, hillslope, and fluvial processes.
Numerical modeling
Numerical modeling is the practice of mathematically describing the physics of a system in a computer program. We can employ numerical models to investigate the general behavior of a physical system (e.g., how will rates of sediment transport change if annual rainfall increases?), diagnose the current state of system that is difficult to observe (e.g., where does our groundwater come from?), and/or predict a system's future evolution (e.g., how many centimeters of sea level rise should Miami residents expect by 2100?).
Field experiments
Remote sensing allows investigation of earth surface processes over huge spatial scales, but is often low resolution in either space (i.e., large pixels), time (i.e., rare image acquisition), or both.
Cross-discplinary science
Though humans define one field of study within certain confines, Nature does not. Aspects within the earth system - climate, surface processes, and tectonics, for instance - influence one another in two-way connections. Further, earth systems affect and are affected by biotic and human societal systems. I am very interested in science that spans these imagined divides.
Figure demonstrating the many link between earth and biotic systems in a glacerized basin. Figure modified from O'Neel et al. [2015]
We see objects as different colors because they reflect and absorb certain wavelengths of light more strongly than others. Remote sensing scientists can exploit this feature to distinguish objects based on their spectral characteristics. Fresh snow appears bright white to us because it is highly reflective to all wavelengths of visible light. Figure created using USGS data.
Undergraduate assistant Will van Gelder measures how well supraglacial stream water transmits electricity, which can be used to calculate water discharge.