Carbon Cycling in Arctic Waters and Response to Climate Change
Photo: M. Fortier
In our current climatic and planetary oceanographic regime, the Arctic Ocean appears to be a net sink of atmospheric carbon dioxide. However, the mechanisms controlling that sink are highly regional, with different biological and physical processes dominating in every new study area. Therefore, in order to understand not only how the total polar carbon sink functions now, but how it might respond to climate change, we have to identify the myriad processes controlling carbon export in a wide range of polar regions, from remote deep waters to shallow coastal seas with their countless influences (see, for example, Miller and DiTullio, 2007). Fortunately, there are a lot of us working on it.
FoxSIPP
Foxe Basin is one of only 3 places in Canada where deep water forms consistently. However, unlike the Labrador and Beaufort Seas, where young deep waters are diluted by mixing into large basins, Foxe Basin is relatively enclosed, with only one deep channel outlet. We've deployed a bottom mooring, fully kitted out with CO2 system sensors, in the deep outlet channel from Foxe Basin to Hudson Strait. As long as we get the mooring and all its sensors back, we should be able to learn something about the relationships between sea-ice formation, deepwater formation, and carbon export. This is one of our many thrilling collaborations with our friends at the University of Calgary.
​
ArcticNet
An on-going collaboration with Brent Else of the University of Calgary, Dave Capelle of the Freshwater Institute, and Tim Papakyriakou of the University of Manitoba is exploring the interannual variability in air-sea CO2 exchange throughout the Canadian Arctic. We have confirmed that at least in the Canadian Archipelago, summer stratification and warming severely limit atmospheric CO2 drawdown, and can even lead to outgassing late in the summer. In addition, UofM Ph.D. student Tonya Burgers has discovered that large-scale circulation of the central Arctic basin can intermittently deliver Russian river waters to the Canadian Archipelago, where they can limit CO2 drawdown.
​
​Arctic Ocean Observatories
In yet another a collaboration with Brent Else of the University of Calgary, we are developing an integrated atmospheric-oceanic observatory in the Canadian Archipelago. The observatory includes a permanent meteorological tower on a small island to measure air-sea (or air-sea ice) CO2 fluxes, and a network of oceanographic stations in the surrounding area to monitor the marine CO2 system. Our ultimate goal is to deploy in-situ pH and pCO2 sensors on an underwater platform upwind of the flux tower, but year-round underwater moorings in such a dynamic ice environment have presented daunting logistical challenges. Thanks to graduate student Patrick Duke and a cabled (i.e., getting electricity from shore) observatory operated by Ocean Networks Canada in Cambridge Bay, we have been able to validate the performance of CO2 system sensors in cold waters with dramatic salinity variations (and identify some of the mechanical things that can go wrong when the sensors are exposed to below-0 °C temperatures for long periods of time). Now that we understand the behaviour of the sensors in cold waters, we can confidently deploy them in more challenging areas where the resulting high-temporal-resolution data will help us understand the full annual carbon cycle across more of the Arctic Ocean.
​
BaySYS
Hudson Bay is an interior Canadian Arctic sea, where the complex mixing between sea-ice melt and river runoff, with their contradictory impacts on the net CO2 sink, results in strong horizontal and vertical pCO2 gradients. In an intensive 2-year study combining carbonate system measurements with hydrographic and O-18 data, we have attempted to understand how the changing hydrological cycles of the Arctic will ultimately feed back onto the climate. University of Calgary graduate student Mohamed Ahmed confirmed that during the spring melt period, the surface pCO2 is dominated by sea-ice melt, and supersaturation due to river waters is limited to the areas around major river mouths. This may be at least partly due to dams on many of the rivers that have spread their flow throughout the year, at the expense of a major spring freshet. Over longer time-scales (i.e., years, instead of seasons), however, the rivers do appear to substantially limit CO2 drawdown; University of Manitoba post-doc Dave Capelle found that remineralization of terrestrial organic matter increases surface pCO2 in the bay by more than 100 μatm. Net result: Hudson Bay appears to be a smaller CO2 sink than other Arctic coastal seas, and that sink may be decreasing with sea-ice loss and increasing river discharge.
​
The Circumpolar Flaw Lead Study (CFL)
This International Polar Year Program involved an overwintering in the Beaufort sea from 2007-2008. My role was primarly consultative, but those who actually did the work did make some very interesting discoveries. Namely, Brent Else found that a broken ice cover might enhance air-sea gas exchange, presumably through some sort of funky hydrodynamical effects. Also, like the Greenland Sea (see below), Amundsen Gulf appears to remain undersaturated in CO2 throughout the winter, making the area a substantial sink.
​
The Canadian Arctic Shelf Exchange Study (CASES)
The CASES project ran from the fall of 2002 until the fall of 2004, and included an overwintering in the landfast ice of the southern Beaufort Sea during 2003-04. I collaborated with Al Mucci of McGill University and Tim Papakyriakou of the University Manitoba to map the inorganic carbon system on the Mackenzie shelf and Amundsen Gulf in both space and time. This allowed us to quantify air-sea CO2 fluxes (a net drawdown, incidentally, at least during the open water season).
​
The International North Water Polynya Study (NOW)
The inorganic carbon team on the NOW project included Tish Yager and Doug Wallace's groups, as well as myself, and based on a comprehensive data set covering the period from April to October, we found that the surface waters were undersaturated in CO2 during the entire ice-free period. These results supported Tish's hypothesis that seasonal ice cover could result in a net atmospheric CO2 sink by limiting outgassing when the surface waters are supersaturated in winter but allowing drawdown during the productive, ice-free summer season. However, we have since discovered that sea ice doesn’t actually limit winter air-sea exchange so much, afterall....
​
The European Subpolar Ocean Programme (ESOP)
The mid-1990s were a period when the rate of deepwater formation in the central Greenland Sea was at a minimum, and from 1993 through 1995, the ESOP1 program conducted a comprehensive, seasonal study of the relationships between that slow deepwater formation rate, sea ice formation, and the carbon cycle. Working with Tom Noji and Francisco Rey (of the Norwegian Institute of Marine Research) and Ingunn Skjelvan of the University of Bergen, we found that although primary production was high and the community was consistent with high f-ratios, biogenic export was extremely low, because deep winter mixing brought remineralized carbon back to the surface from as deep as 800 m. Nonetheless, the Greenland Sea was still a sink of atmospheric CO2 throughout the entire year, thanks to low temperatures and long-term solubility pump activity.
​
A Long Time Series...
A data set from the Beaufort Sea, going back into the 1970s and rescued from obscurity by Karina Giesbrecht, has confirmed that pCO2 is increasing and pH is decreasing in the sub-surface waters. This trend is a result of changes in CO2 content and alkalinity, not temperature or salinity. That is, these waters are becoming more corrosive to carbonate minerals because of anthropogenic carbon inputs, possibly coupled with ecosystem changes, rather than natural reorganizations of water masses.
​
