Sub-seasonal Arctic Lake Area Variability

Arctic and Boreal regions are home to the highest density/number of lakes on Earth. Understanding how these lakes are changing, on both seasonal and interannual time scales, is critical for assessing environmental sensitivity to climate change, monitoring the hydrology of these remote and often poorly studied regions, and constraining terrestrial freshwater methane emissions. While many studies have examined long-term, interannual changes in lake area, until recently, little research had focused on shorter-term changes in lake area due to a traditional tradeoff between high spatial and high temporal resolution satellite imagery.

 Over the past 5 years, there has been rapid development of new, tiny satellites known as CubeSats by companies such as Planet Labs which now provide imagery of the entire land surface of the Earth at 3 meter resolution, every single day. This imagery has opened new possibilities for mapping short-term changes in lake area, but due to the newness of the CubeSat technology, it first required testing and development of novel methodology. The first project of my PhD involved testing a water classification and lake area change tracking method using Planet CubeSat imagery over a small area of the Yukon Flats during summer 2016. In this paper, published in Remote Sensing, we demonstrated that CubeSat imagery could be used to accurately map water area and thus could provide valuable insight into seasonal wetland and lake dynamics.

Animation of changes in lake area observed using CubeSat imagery.

Animation of changes in lake area observed using CubeSat imagery.

Following on this project, I then upscaled this method to be applicable over a much larger area by incorporating machine learning. As part of NASA’s Arctic-Boreal Vulnerability Experiment, I tracked changes in lake area over summer 2017 in four study areas in Northern Canada and Alaska, corresponding to the locations of our summer 2017 fieldwork. We identified a surprisingly hydrologically dynamic landscape, particularly in the poorly studied Canadian Shield where we found substantial sub-seasonal lake area variability. These results, published in Geophysical Research Letters, have implications for the estimation of freshwater methane and carbon dioxide emission, which is often higher along seasonally fluctuating lake margins. The press release for this paper can be found here.

Gridded maps of lake dynamism (seasonal maximum minus seasonal minimum area, km2) for the four study areas analyzed in the 2019 GRL paper. White dashed lines show major river systems, and white stars indicate locations of in situ lake level measurements.

Gridded maps of lake dynamism (seasonal maximum minus seasonal minimum area, km2) for the four study areas analyzed in the 2019 GRL paper. White dashed lines show major river systems, and white stars indicate locations of in situ lake level measurements.


Related Publications

Cooley, S.W., L.C. Smith, J.C. Ryan, L.H. Pitcher and T.M. Pavelsky (2019), Arctic-Boreal lake dynamics revealed using CubeSat imagery, Geophysical Research Letters. doi.org/10.1029/2018GL081584

Pitcher, L., T. Pavelsky, L. Smith, D. Moller, E. Altenau, G. Allen, C. Lion, D. Butman, S. Cooley, J. Fayne, and M. Bertram (2018), AirSWOT InSAR mapping of surface water elevations and hydraulic gradients across the Yukon Flats Basin, Alaska, Water Resources Research. doi.org/10.1029/2018WR023274

Cooley, S.W., L.C. Smith, L. Stepan and J. Mascaro (2017), Tracking dynamic northern surface water changes with high frequency Planet CubeSat imagery, Remote Sensing 9, 1309. doi.org/10.3390/rs9121306