The Earth's poles provide the most direct record of recent climate change and also are where massive ice reservoirs are vulnerable to melting. I conduct research at both poles to determine how polar surfaces are changing on annual and decadal timescales. I integrate these field data with orbital observations to develop thorough reconstructions of polar surface evolution.

* * *

Absorption of atmospheric water by salts on the ground (deliquescence) is one way that the polar desert of Antarctica is capable of making small amounts of liquid brine. This process is extremely difficult to document but we used long-duration (several months) high-frequency (every 5 minutes) time-lapse photography combined with in-situ measurements of relative humidity to fully describe how this process behaves on a daily cycle. As observed in the video above in Garwood Valley in the McMurdo Dry Valleys, bright salts are observed in daytime but turn dark when they generate brines in the evening when the valley falls under shadow and the relative humidity rises.

This fills in critical temporal gaps in our observations of Antarctica from orbit, which cannot be achieved at this rapid interval, and provides clues about whether similar processes could be generating brines on Mars.

For more information, see Ariel Deutsch's paper fully detailing this process. Deutsch et al., 2022.

* * *

The McMurdo Dry Valleys of Antarctica receive less than 10 cm of percipitation per year, and most of that is in the form of snow that sublimates rapidly. Yet stream channels still form from meltwater on an annual basis. How is this possible? In some areas, massive glaciers provide a perennial source of ice to melt, but what about regions without perennial water sources?

We focused on the South Fork of Upper Wright Valley, where stream channels are active but no large glaciers are observed. We used long-duration, high-frequency time-lapse imaging to fully characterize a sprawling surface and near-surface drainage system that provides meltwater that eventually travels all the way to Don Juan Pond at the lowest point of the valley.

Below, notice how time-lapse imaging revealed near-surface dendritic drainage networks after a snowstorm. Soil with high moisture content had a higher heat capacity than neighboring dry soils, so snow melted/sublimated faster, thus revealing these previously invisible conduits.



Within the channels themselves, melting is highly variable from season to season. Our results indicate that massive flood events like the one observed below are controlled not by where ice is most likely to melt, but by where it is most likely to accumulate during Antarctic winters, goverened by topographic shading and environmental conditions.



For a full analysis of the behavior of Antarctic non-glacier-fed streams over a full decade, see Dickson et al., 2017.

* * *

Time-lapse imaging during austral winter allows for unique observations of how the surface of Antarctica changes when field work is not possible. Our cameras have worked through the frigid Antarctic winter and captured light reflected off of the gibbous Moon, allowing for the assessment of surface snow deposits and lake-level measurements of Don Juan Pond.

Commonwealth Stream in June, 2015. Illumination is entirely from light reflected off of the full moon in the upper-left of the image.

For more visualizations of Antarctica under moonlight, please visit here.

* * *

The McMurdo Dry Valleys, Antarctica, provide the most pristine landscape on Earth to investigate how small changes in climate result in landscape change. This surface response has been historically challenging to document in the Dry Valleys due to extremely low rates of change, which are typically too low to be perceptible with traditional field observations.

Time-lapse imaging reveals this surface response at unprecedented spatial and temporal resolution. Cameras record images every 5-minutes for several months during peak austral summer, when conditions at the surface briefly surpass the triple point of H₂O. These images are synchronized with meteorological data to close the climate forcing/surface response loop, and allow for the testing of detailed hypotheses regarding Antarctic landscape evolution.

See summary as reported in Dickson et al., 2014.

* * *

Brine generation in Don Juan basin, Antarctica. Time-lapse data show water tracks hydrating at the exact moment that a front of moist air passes through Upper Wright Valley. This is confirmation that salts (specifically CaCl2) absorb water out of the atmosphere, generating brines that match the composition of Don Juan Pond, the saltiest body of water in the world. From Dickson et al., 2013.

* * *

* * *

Melting of the Garwood Ice Cliff, Antarctica. A stranded ice mass is exposed on its equator-facing side. Time-lapse imaging has shown that its rapid recession (~1 m/yr) is due to direct radiation, as opposed to undercutting by the Garwood River. From Levy et al., 2013.

---

I am currently part of a large NSF-funded project that monitors river bank erosion in arctic Alaska in an effort to aid native communities who are at risk from increased melting. I lead the design and construction of time-lapse monitoring stations that document this process in high spatial and temporal resolution, and I wrote the software to correlate surface activity with met station data to determine the climatic force responsible. This is brand new research that will be continued through 2026.