Remote Sensing of Arctic Landscape Dynamics

Abstract

Amplified warming in the Arctic has likely increased the rate of landscape change and

disturbances in northern high latitude regions. Remote sensing provides a valuable tool

for assessing the spatial and temporal patterns associated with arctic landscape dynamics

over annual, decadal, and centennial time scales. In this dissertation, I focused on remote

sensing studies associated with four primary components of arctic landscape change and

disturbance: (1) permafrost coastline erosion, (2) thermokarst lake dynamics, (3) tundra

fires, and (4) using repeat airborne LiDAR for the measurement of vertical deformation

in an arctic coastal lowland landscape. By combining observations from several high

resolution satellite images for a 9 km segment of the Beaufort Sea Coast between 2008

and 2012, I demonstrated that the report of heightened erosion at the beginning of the

2000s was equaled or exceeded in every year except 2010 and that the mean annual

erosion rate was tightly coupled to the number of open water days and the number of

storms. By combining historical aerial photographs from the 1950s and 1980s with

recent high-resolution satellite imagery from the mid-2000s, I assessed the expansion and

drainage of thermokarst lakes on the northern Seward Peninsula. I found that more than

half of the lakes in the study area were expanding as a result of permafrost degradation

along their margins but that the rate of expansion was fairly consistent (0.35 and 0.39

m/yr) between the 1950s and 1980s and 1980s and mid-2000s, respectively. However, it

appeared that in a number of instances that expansion of lakes led to the lateral drainage

and that over the 55-year study period the total lake area decreased by 24%. While these studies highlight the utility of quantifying disturbance during the remotely sensed image

archive period (~1950s to present) they are inherently limited temporally. Thus, I also

demonstrated techniques in which field studies and remote sensing data could be

combined to extend the identification of landscape disturbance events that occurred prior

to the remote sensing archive. I identified two large regions indicative of past

disturbance caused by tundra fires on the North Slope of Alaska, which doubled the

delineated area of tundra fire disturbance on the North Slope over the last 100 to 130

years. I conclude the dissertation by demonstrating the utility of repeat airborne light

detection and ranging (LiDAR) data for arctic landscape change studies, in particular

vertical surface deformation, and provide momentum for going forward with this

emerging technology for remote sensing of arctic landscape dynamics. The

quantification of arctic landscape dynamics during and prior to the remote sensing

archive is important for ongoing monitoring and modeling efforts of the positive and

negative feedbacks associated with amplified Arctic climate change.