Electronic Thesis and Dissertation Repository

Thesis Format

Integrated Article


Master of Science



Collaborative Specialization

Planetary Science and Exploration


Osinski, Gordon R.


Permafrost underlies 50% of Canada’s terrain and underlies 24% of the Earth’s total land area. It is a major driving force in the generation and evolution of patterned ground landforms such as polygons, stone circles, mud boils, and stripes, etc. that are seen on both the surface of the Earth and Mars, specifically in periglacial domains. The distribution of subsurface ice in these landforms (i.e. polygonal terrain) on Earth is a key constraint on past climate and process-form relationships in high arctic and periglacial regions. These landforms also have the potential of storing ice in the subsurface meaning that the volumetric concentration of buried ice can provide information about the processes that led to its deposition. In this research, two Canadian High Arctic study sites were analyzed: Strand Fiord, Axel Heiberg Island (ᐅᒥᖕᒪᑦ ᓄᓈᑦ, Umingmat Nunaat) and Haughton River Valley, Devon Island (ᑕᓪᓗᕈᑎᑦ, Tallurutit). Methods in this research involved the utilization of (1) ground penetrating radar (GPR) for 3D ice-wedge volume estimation, (2) Light Detection and Ranging (LiDAR) for hyper-resolution digital elevation model (DEM) generation and quantifying the degree of spatial sorting in patterned ground using Kernel Density Estimation (KDE), (3) sedimentology for grain size and facies association in order to infer depositional environment, (4) fieldwork notes and measurement, and (5) drone and aerial photos for 3D photogrammetry models. In Strand Fiord, ice-wedge geometry and volume were calculated to be asymmetric and 43.28 m3 respectively in a 25 m x 25 m grid within the study area; ice-wedge polygons were associated with areas of active transition (i.e. secondary/tertiary polygons that have experienced multiple episodes of cracking). In the Haughton River Valley, patterned ground spatial distribution and sorting were related to the local periglacial region such as microtopography, slope and lithology wherein fine (clay, sand) sediments were often associated with the formation of frost-shattered nets, polygons, and mud boils in plateau tops, while the coarse (gravel, cobble) sediments were associated with stone circles and stripes on the valley floor/slopes. This research establishes the constituents of a periglacial landsystem and its significance as a foundation for the development of more comprehensive models that may be used to enhance understanding of the process-form relationships in the periglacial Arctic and further, ice-rich planetary systems.

Summary for Lay Audience

The Canadian High Arctic Archipelago has many similarities to ice-rich planetary bodies like Mars. On Earth, permafrost (or frozen ground) in high-latitude and high-altitude areas is the main driving factor for changes in the landscape overtime. As years pass, permafrost terrain that overlay massive volumes of ice underground start to degrade and shape landforms such as patterned ground. Examples of periglacial patterned ground landforms include honeycomb like-polygons, stone circles, mud boils, gully-polygons (or “gullygons”) and more. Surprisingly, these landforms can also be seen on Mars, which poses the question: Are we able to estimate how much ice there is underneath the surface of Mars by looking at the landforms on Earth? Will investigations on Earth be able to infer how the landscape of Mars looked/will look like in the past/in the future? This research goes into depth on how these potentially ice-rich landforms are all connected and affect the formation of one another using different fieldwork and mapping methods – a concept called landsystem analysis.

Fieldwork was done in Devon Island, Nunavut in July 2018/July-August 2019 as well as Axel Heiberg Island, Nunavut in July 2019. This was done through three different methods: (1) sedimentology; measuring sediment grains, (2) geomorphology; mapping landforms, (3) ground penetrating radar (GPR); a non-invasive towed system to classify different soil layers, and (4) kinematic light detection altimetry and ranging; a laser on a backpack for 3D modelling of the terrain. Data collected showed that ice-wedge polygons often occur in secondary and tertiary polygons (multiple episodes of cracking) locally in the middle of transition areas within the landscape (i.e. smaller to bigger polygons, from the mouth of an alluvial fan to the terminus; Chapter 2). It also showed that 3D-GPR is a powerful tool to visualize the ground for the shape/size of ice-wedges in order to estimate ice volume (Chapter 2). In combination with quantitative geomorphology (Chapter 3), we were also able to measure the compactness, elongation ratio, and other shape descriptors of patterned ground and found a potential trend.

All in all, periglacial (permafrost) landforms and patterns can provide a better understanding of how buried ice behaves and how the landscape of the Canadian High Arctic Archipelago evolves over time as a resource for communities/planetary exploration.