Authors: Yuhong He*, University of Toronto Mississauga, Mitchell Bonney, University of Toronto Mississauga
Topics: Remote Sensing, Landscape, Development
Keywords: Landsat, Vegetation Change, Urban-rural Landscapes, Post-development
Session Type: Virtual Paper
Start / End Time: 8:00 AM / 9:15 AM
Room: Virtual 33
Presentation File: No File Uploaded
The Landsat archive, with nearly 50 years of Earth observation data across the planet, represents the most long-term, consistent, and standardized remote sensing view of landscape change. Recent efforts in standardizing imagery across all sensors and years (e.g., LandsatLinkr), fitting models to noisy reflectance data (e.g., LandTrendr, Continuous Change Detection), and having all these data and analysis techniques available in the cloud through Google Earth Engine provides many new opportunities in landscape change modeling. In this study, we test the capability of the full Landsat record (1972-) for quantifying development and post-development vegetation change across rural, suburban, and urban landscapes in the Region of Peel (west of Toronto in southern Ontario). Initial results are most promising in suburban settings, where percent canopy cover change can be estimated through time and used to understand how these communities are greening in the decades post-development. Different post-development municipalities are greening at different rates, representing the possible influence of socio-economic factors on urban forest change. Overall, the extension of the standardized Landsat archive to 1972 allows for the observation of much suburban development that would have otherwise been missed. Furthermore, the development of this workflow in Google Earth Engine will allow for the extension of the results to other cities and a much broader understanding of how urban forests and vegetation are changing in the decades post-development.