Methane Fluxes in the High Northern Latitudes for 2005 - 2013 Estimated Using a Bayesian Atmospheric Inversion

In recent decades, the high northern latitudes have warmed substantially with temperatures in the Arctic increasing at an average rate of 0.38°C per decade. This warming may have a considerable impact on methane (CH4) emissions, especially from wetlands. In addition, there are important anthropogenic emissions, particularly from oil and gas, in the high northern latitudes. Given the rapid rate of warming and possible future expansion of anthropogenic activities in the Arctic, it is imperative to have accurate estimates of present-day CH4 emissions, to better understand their natural variability, and to determine if there is any trend. We estimated CH4 fluxes for 2005 to 2013 for the high northern latitudes (north of 50°N) using a Bayesian inversion. Our inversion is based on atmospheric transport modelled by the Lagrangian particle dispersion model, FLEXPART, and CH4 observations from 17 in-situ and 5 flask-sampling sites distributed over northern North America and Eurasia. CH4 fluxes were determined at monthly temporal resolution and on a variable grid with maximum resolution of 1° × 1°. For northern North America, we found a mean source of 16.6 - 17.9 Tg y-1, which is dominated by fluxes in the Hudson Bay Lowlands (HBL) and western Canada, specifically, the province of Alberta. We find a source for HBL of 2.7 - 3.4 Tg y-1, which is close to land surface model estimates of wetland fluxes and to other independent inversion estimates. For Alberta, we find a source of 5.0 - 5.8 Tg y-1, which is significantly higher than anthropogenic and wetland emission estimates from inventories and land surface models. Furthermore, this source is persistent throughout winter suggesting that the emissions are mostly of anthropogenic origin. For Northern Eurasia, we find a mean source of 52.2 - 55.5 Tg y-1, with a strong contribution from fluxes in the Western Siberian Lowlands (WSL) for which we estimate a source of 19.3 - 19.9 Tg y-1. Over the 9-year inversion, we find significant year-to-year variations in the fluxes. In northern North America, particularly in HBL, these appear to be driven by changes in soil temperature, while in WSL, these appear to be more driven by soil moisture.