Detecting Changes in Arctic Carbon Fluxes

Warming in the Arctic has proceeded faster than anyplace on Earth. Our current understanding of biogeochemistry suggests that we can expect feedbacks between climate and carbon in the Arctic. Changes in terrestrial fluxes of carbon can be expected as the Arctic warms, and the vast stores of organic carbon frozen in Arctic soils could be mobilized to the atmosphere, with possible significant impacts on global climate. Quantifying trends in Arctic carbon exchanges is important for policymaking because greater reductions in anthropogenic emissions may be required to meet climate goals.

Observations of greenhouse gases in the Arctic and globally have been collected for several decades. Analysis of this data does not currently support significantly changed Arctic emissions of CH₄, however it is difficult to detect changes in Arctic emissions because of transport from lower latitudes and large inter-annual variability. Unfortunately, current space-based remote sensing systems have limitations at Arctic latitudes. Modeling systems can help untangle the Arctic budget of greenhouse gases, but they are dependent on underlying prior fluxes, wetland distributions and global anthropogenic emissions. Also, atmospheric transport models may have significant biases and errors. For example, unrealistic near-surface stability can lead to underestimation of emissions in atmospheric inversions.

We discuss our current understanding of the Arctic carbon budget from both top-down and bottom-up approaches. We show that current atmospheric inversions agree well on the CH₄ budget. On the other hand, bottom-up models vary widely in their predictions of natural emissions, with some models predicting emissions too large to be accommodated by the budget implied by global observations. Large emissions from the shallow Arctic ocean are also inconsistent with atmospheric observations. We also discuss the sensitivity of the current atmospheric network to what is likely small, gradual increases in emissions over time by examining modeled and observed spatial and seasonal variability. An issue we will consider is whether well-mixed background atmospheric records are more likely to detect changing Arctic emissions compared to stronger, but more variable signal from local sources.