Comparative Analysis of Estimates of Carbon Emissions from Wildfires in Russia Based on Global Remote Sensing Products

This paper presents wildfire carbon emissions estimates in Russia, 1997–2023, derived from global satellite products. The following products developed using Earth remote sensing (ERS) data were included in the review: GFASv1.2, GFEDv4.1s, FEERv1.0-GFASv1.2, FINNv2.5, and QFEDv2.5-r1. Depending on the product, estimated 130–275 TgC, or 6–11% of global wildfire carbon emissions originate from the territory of Russia (3.5–7.5% considering forest fires only). Annual wildfire carbon emissions range from 54 to 490 TgC depending on the product and annual fire activity, with maximum emissions in years of extremely large fires (2003, 2008, 2012, 2021). Forest fires account for 53–72% of total pyrogenic carbon emissions in Russia. For most of the products reviewed, emissions estimates exceed official wildfire reporting data, which only consider fires in managed forests (long-term average 80 TgC). Over the observation period 2002–2023, an uptick in forest fires carbon emissions (+0.5 TgC/year) was observed, in contrast to a reduction in emissions from wildfires in nonforested areas (–1.25 TgC/year). Annual pyrogenic carbon emissions estimates show a high correlation (R2 > 0.8) between the reviewed products. However, on a pixel-wise level (spatial resolution 0.1°), the correlation between the products' estimates is significantly lower (R2 = 0.22–0.44). For the products in review, it is estimated that 0.5–6.6% of emissions come from non-wildfire sources, such as gas flares in oil and gas facilities.