Terrain analysis, snow-depth interpolation, and remote sensing workflows supporting wildfire severity modelling in Arctic tundra ecosystems.

In this team-based remote sensing project, I led the terrain analysis and snow-depth modelling workflow, including DEM mosaicking, slope analysis, snow interpolation, and spatial integration of environmental variables used in wildfire severity modelling. The project combined ArcticDEM terrain products, MODIS snow phenology, interpolated Yukon snow survey data with satellite-derived burn severity metrics, NDVI and soil moisture analysis to create a model of pre-season drivers of wildfire severity in Northern Yukon tundra systems.

<aside> 👩🏻‍💻

Role

Terrain analysis & snow modelling

</aside>

<aside> 🛰️

Data

ArcticDEM, MODIS, Landsat, Sentinel, Yukon snow survey

</aside>

<aside> ⚒️

Tools

QGIS, Google Earth Engine, R

</aside>

<aside> 🗺️

Methods

DEM mosaicking, slope analysis, IDW interpolation, snow phenology

</aside>

Project Context

Wildfire severity is increasing across Arctic and sub-Arctic ecosystems as climate warming alters vegetation structure, moisture availability, and snow dynamics. Understanding how terrain and pre-season snow conditions influence fire severity is increasingly important for ecosystem resilience and habitat conservation in Northern Yukon tundra systems.

This project examined how elevation, snow depth, precipitation, soil moisture, and vegetation greenness influenced wildfire severity across five Northern Yukon fire sites between 2017–2022.

My Contributions

Terrain Analysis

• Processed and mosaicked ArcticDEM elevation tiles
• Generated DEM and derived slope surfaces in QGIS
• Compared terrain gradients against snow distribution patterns
• Produced cartographic terrain visualizations for analysis and presentation

Snowpack Modelling

• Compiled Yukon Government snow survey station data
• Designed and executed inverse-distance weighted (IDW) snow-depth interpolation workflows
• Created multi-season snow-depth surfaces at 250 m resolution
• Integrated MODIS-derived snow cover duration metrics into broader environmental analysis

Geospatial Integration

• Helped integrate terrain and snow variables into wildfire severity modelling workflows
• Supported multi-source raster harmonization and spatial comparison across resolutions
• Created the GitHub repository and designed the workflow for team contributions

Study Area

The study focused on five wildfire sites in Northern Yukon tundra and taiga environments, including areas overlapping Vuntut National Park and Porcupine Caribou Herd habitat.

Terrain Analysis & Snow Depth Interpolation Workflow

Snow-depth observations from three Yukon Government monitoring stations were interpolated across the study area using inverse-distance weighting (IDW) to estimate regional snowpack distribution.

Higher snow-depth regions aligned broadly with lower-elevation southern terrain zones, while steeper mountainous terrain showed reduced snow accumulation patterns.