The size and spatial distribution of organic carbon stocks in terrestrial ecosystems of Canada

Abstract:

With 40% of its territory in the arctic circumpolar region, Canada’s terrestrial ecosystems store the second largest soil organic carbon (SOC) stock of the world. As temperatures rise, SOC is becoming available for decomposition and eventual release into the atmosphere, which makes the quantification of SOC stock of Canada of high interest for the assessment of climate change impacts. Therefore, this study aims to estimate the total (Pg C) and spatially explicit SOC maps with 250 m spatial resolution and three depth intervals (0-1m, 0-2m and 0-4m) in Canada.

We used 6,530 ground soil samples, long-term climate data, remote sensing observations and a machine-learning method to model and generate spatial distribution of SOC contents for entire Canada. The soil samples (observational variable) containing the x and y coordinates, depth and SOC information were overlaid with the stacked covariates (soil forming factors) to compose the regression matrix. Random forest models were trained using a recursive feature elimination scheme and a cross-validation assessment. The best model was used for spatial prediction of SOC over Canada in intermediate depths between 0 and 4 m. Afterwards, the SOC content maps were corrected with bulk density (BD) and coarse fragment (CF) information to compute the total carbon stock for each horizon. The horizons have been added to compose the three depth intervals multiplied by root depths (RD) fraction. Ultimately, the SOC stock estimates were multiplied by area to provide the total SOC in Pg C.

Results showed that RF model associated with 25 selected covariates was successful in capturing the major variation in SOC content across the country, with an R² of 0.83. Depth was the most important covariate for predicting SOC, followed by DEM, climate features (long-term mean temperature and precipitation), and finally vegetation indices. The total SOC calculated for Canada was 316 Pg (37 kg/m²), 526 Pg (49 kg/m²) and 1,547 Pg (182 kg/m²) in 0-1m, 0-2m and 0-4m depths, respectively. In these depth intervals, Hudson Plain ecozone alone stores approximately 28 Pg C, 52 Pg C and 113 Pg of C, respectively. In order to estimate the total SOC of Canada’s terrestrial ecosystems, we are also estimating the C stock of vegetation using SAR, LiDAR and reflectance data from multiple satellite sensors.