Abstract
Light detection and ranging (LiDAR) technologies are changing the ways
in which scientists research the Arctic. Unmanned aerial vehicle
(UAV)-based LiDAR collects detailed structural landscape data by
returning high density point clouds. LiDAR systems are improving the
quality and accuracy of data collection compared to field surveys and
help to remove some of the logistical barriers of research in remote and
complicated terrain. Our study mapped thermokarst depressions in a 3 km2
watershed on the Seward Peninsula near Nome, Alaska in 2017 and 2018.
The watershed is characterized as tussock permafrost landscape
consisting of grasses and mosses interspersed with patches of dense
shrubs. By configuring the UAV with a 32 laser swath and flying slowly
at low altitude, we collected high density point clouds of about 4,000
points m2, including high density terrain surface points underneath
dense shrubby vegetation. We then modeled the sub-vegetation terrain
surface at very fine detail to detect thermokarst depressions. Combining
these high resolution data with vegetation surveys and topographic
properties, we tested the relationship between permafrost subsidence,
thermokarst depressions and vegetation type, specifically the
relationships in shrub-associated thermokarst features. By coupling our
LiDAR data and analysis with hydrologic models, climate variables (e.g.,
snow depth, soil moisture), and vegetation surveys, we can infer
geospatial relationships between thermokarst development, vegetation,
and landscape position throughout the watershed. The technologies used
in our study have implications for predicting the development of future
thermokarst features and permafrost thaw sites across the Arctic.