Working Group 3: What are the requirements for detecting a climate change signal in the North Pacific carbon cycle?

Working groups 1 and 2 are focused on identifying changes in air-sea flux and within the water column, respectively. Our working group, WG-3, is intended to be a cross-cutting group aimed at trying to determine how and when we might be able to distinguish a long-term climate change signal from natural seasonal and interannual variability. There have been many papers published with relatively long-term (2-20 year) surface ocean pCO₂ records that have noted a certain amount of seasonal and interannual variability in the records, but then fit these data with a trend line and attribute that to long-term change. Is this correct or are there other issues, like PDO type shifts, that need to be considered? Within the water column we have approaches that we believe allow us to isolate the anthropogenic CO₂ component from the measured dissolved inorganic carbon (DIC) value relative to a preindustrial ocean. Thus, we believe that we can identify long-term changes over a 200 year timeframe. These approaches, however, rely on assumptions like steady state circulation and biology that probably are not valid at some level. When we compare two or more cruises separated by a few years we also see changes. How can we tell whether these are long-term climate change or part of the natural variability of the ocean?

Andy Jacobson has started us off by sharing quite a provocative result showing decreasing PP in CZCS versus SeaWiFS, and increasing PP in the HadCM3L model. What trends do the rest of you see in your models? Do any of you have any insights on whether we can even trust the CZCS versus SeaWiFS comparison (Francisco or Ragu?)? Andy, do you have any insight into why the PP went up in the North Pacific in the model? It is very interesting that it does.

The above focuses on PP. What other things have any of you looked at that we ought to be thinking about? Can any of you summarize Luanne Thompson's latest modeling efforts at UW (Paul? Jim?)? Keith, what do you see in the NCAR model? Are there any significant shifts in the functional groups we should know about? Holger, have you guys done some predictions with analyzed winds that we ought to know about? Ragu, how about your group? And what do the observations tell us about what is happening (Francisco, Paul, Jim, anyone else?)?

I would also much appreciate hearing from Ono and Andreev with a summary of work they have been doing in this area, as I am not familiar with their work. Indeed, perhaps it would be very helpful if each of us could send to the rest a list of one or two papers we may have written on this topic that we may want to look at ahead of time (perhaps we can begin to put together a bibliography ahead of time), as well as a list of topics you think we should talk about at the meeting and discuss in our paper. Clearly primary production is one of the important topics we should look at. Jim and others, what thoughts do you have about what topics are important for us to think about?

Please respond to everyone in the group so we can get a dialogue going!