Canadian Forest Service Publications
Disturbance-informed annual land cover classification maps of Canada’s forested ecosystems for a 29-year Landsat time series. 2018. Hermosilla, T., Wulder, M.A., White, J.C., Coops, N.C., Hobart, G.W. Canadian Journal of Remote Sensing, 44:1, 67-87.
Issued by: Pacific Forestry Centre
Catalog ID: 39135
Availability: PDF (download)
Available from the Journal's Web site. †
† This site may require a fee
Land cover classification of large geographic areas over multiple decades at an annual time step is now possible based upon free and open access to the Landsat data archive. Annual gap-free, best-available-pixel, surface reflectance, image composites and annual forest change maps have been generated for Canada for the years 1984 to 2012. Using these data, we demonstrate the Virtual Land Cover Engine (VLCE), a framework for change-informed annual land cover mapping, over the 650 million ha forested ecosystems of Canada, to produce a 29-year data cube of land cover. Post-processing aimed to reduce spurious class transitions is undertaken integrating change information, land cover transition likelihoods, and year-on-year class membership likelihoods. Validation was assessed for a single year (2005) using independent data for an overall accuracy of 70.3% (± 2.5%). Key results are the detailed capture of trends in land cover, illustration of land cover links to disturbance processes, and insights related to the general stability of land cover over time with stand replacing disturbance followed by regeneration of forests. The portable mapping framework and resultant data products offer an integrated, long baseline, disturbance-informed and detailed depiction of land cover to meet science and program related information needs.
Plain Language Summary
Land cover and land cover change information are critical for forest monitoring and reporting activities, as well as for applications such as carbon budgets, biodiversity assessments, and habitat characterizations, among others. Land cover classification of large geographic areas over multiple decades at an annual time step is now possible based upon free and open access to the Landsat data archive, combined with advances in algorithms and computational capacity. Annual gap-free best-available-pixel (BAP) surface reflectance image composites from Landsat have been generated for Canada for the years 1984 to 2012 (a tessellation of >12 billion pixels at 30 m spatial resolution). From these image-composite time series, annual forest change information has been derived and attributed to disturbance type. Availing upon the image composites and associated attributed change information, herein we demonstrate the implementation of a change-informed annual land cover mapping framework for the 650 million ha forested ecosystem area of Canada: the Virtual Land Cover Engine (VLCE). Following classification using Landsat optical reflectance channels and terrain derivatives (totaling ~17 Tb), annual class membership likelihood maps are produced for all 29 years (~7 Tb of classified outputs). Time series change information to control for presence of disturbance and class transition likelihoods are then integrated with the annual, class-level likelihoods to produce a 29-year data cube of land cover through processing of over a trillion pixels.