“We never had doubts about our technical capacity to see this project through to the end,” says Camile. “From the beginning, the greatest challenge was finding the data we needed.”
In the beginning, though, the success of the Mapping Canada’s Carbon Landscape study was not a foregone conclusion. Camile Söthe, the postdoctoral fellow in Alemu’s lab who led the study, faced the daunting challenge of compiling and analyzing tens of thousands of field measurements to create a comprehensive map of the carbon stored in Canada’s plants and soils.
“Knowing where carbon is stored in Canada allows us to strategically protect and manage the right places to prevent billions of tonnes of carbon from being released into the atmosphere,” says Megan Leslie, WWF-Canada President and CEO. “Protecting these areas will also benefit wildlife by safeguarding habitat for important species at risk.”
The Remote Sensing Laboratory was a natural partner for WWF-Canada because of its ability to process vast quantities of data — in the case of Canada’s carbon stocks, more than five terabytes worth of data collected from nine million square kilometers of diverse land ecosystems.
To make sense of all this data, Camile and a handful of student assistants used a machine learning algorithm to estimate carbon content based on satellite data. In addition to the final carbon stocks, they produced uncertainty maps that spatially indicate confidence intervals of their estimates.