In May 1999, the Arctic Ocean Science Board announced plans to study the two-way oceanic exchanges that link the Arctic Ocean with subarctic seas (Dickson et al., 1999, AOSB Newsletter 3(2)). The rationale is bound up with the fact that most projections of greenhouse gas induced climate change anticipate a weakening of the thermohaline circulation (THC) in the North Atlantic in response to increased freshening and warming in the subpolar seas (Rahmstrof 1999, Nature 399, 523-524; Rahmstrof and Ganopolski 2000, Climatic Change 43, 353-367; Delworth and Dixon, 2000, J. Clim. 13, 3721-3727). Since the overflow and descent of cold, dense waters across the Greenland-Scotland Ridge is a principal means by which the deep ocean is ventilated and renewed, the suggestion is that a reduction in upper-ocean density at high northern latitudes will weaken the THC. Unfortunately, our models do not yet deal adequately with many of the mechanisms believed to control the THC, and our observations cannot yet supply many of the numbers they need. Though we have a first indication that the coldest, densest part of the Faroe-Shetland overflow may be decreasing (Hansen et al., 2001, Nature 411, 927-930), our present observations of the large scale overturning circulation in the North Atlantic (or anywhere else) are insufficient to detect whether or not it is changing; we have no measurements of the freshwater flux between the Arctic Ocean and Atlantic by either of its two main pathways; we have new measurements (from the EC VEINS project) of the heat and salt flux to the Arctic Ocean but not yet of its variability on any scale; we have a growing knowledge of the long-term variability of dense overflows which “drive” the THC but only embryonic ideas as to their causes, etc. Understandably then, we would take the view that these key mechanisms and processes are too crudely represented in the present generation of climate models. Palaeoclimate records, however, show that massive and abrupt climate change has occurred in the Northern Hemisphere, especially during and just after the last Ice Age (Broecker, 1997, Science 278, 1582-1588; Broecker, 2000, PNAS 97, 1339-1342; Marotzke, 2000, PNAS 97, 1347-1350), with THC change as the most plausible driver, and both palaeoclimate records and models suggest that the changes in the strength of the THC may occur rapidly, in a few decades. Further, in our admittedly-short modern records of ocean variability, we have growing evidence that hydrographic changes of decadal scale in the Arctic and subarctic seas are able to feed south across the deep northern overflows to cause hydrographic changes in the deep and abyssal layers of the Labrador Sea. These variations are large and longsustained - e.g. the freshening of both dense overflows by between 0.01 and 0.02 per decade for the past 3.5 decades (Dickson et al. 2002, Nature 416, 832-837) though we don't yet know enough about process to determine their climatic significance. The high northern latitudes and the ocean fluxes that connect them to adjacent seas are plainly not the only constituent parts of this problem. The THC is driven globally by upwelling, downwelling and a strong component of upper-ocean windforcing, and fluctuations in any one of these components might affect the strength of the THC (see, for example, Toggweiler and Samuels 1995, Deep-Sea Res. 42, 477-500, for the role of the Southern Ocean windfield). Nonetheless buoyancy loss in the northern high latitudes and the factors that control it are still of a fundamental importance, are areas of continuing ignorance and are becoming tractable by modern observing systems. These thermohaline controls and linkages, then, form the research focus of ASOF.
On 6 April 2000 in Cambridge UK, as a joint initiative of the Arctic Ocean Science Board and the International Arctic Science Committee, a discussion meeting on the Sustained Monitoring of Arctic Fluxes was held during Arctic Science Summit Week, with three main objectives. First, to discuss the palaeo- and modelling evidence that THC slowdown or shutdown has happened in the past and is likely to recur in the future. Second, to begin to define the system of critical measurements that will be needed to understand the role of the high-latitude oceans in decadal to centennial climate variability. And third, to discuss ways of achieving the co-ordinated long-term stamina in our funding that we will need if we are to implement such a system across all the main gateways to/from the Arctic Ocean over a period of a decade or more. The scientific planning of ASOF was later advanced by means of a second discussion meeting and workshop, held at the Norsk Polarinstitutt, Tromso on 21-24 September 2000, in conjunction with the H. U. Sverdrup Symposium. Whereas the original meeting had focused on the fundamental questions "Has THC shutdown happened before?" and "Are we right to assume it can recur?" the Tromso workshop had the aim of providing a more complete description of the required observing system, with preliminary costs, and with some results in support (where these exist). The design of an ASOF array was further refined at a National Academy of Sciences Workshop on Abrupt Climate Change: Science and Public Policy, held at Lamont Doherty Earth Observatory, Palisades NY, on October 30-31 2000. Throughout this evolution, discussion has been guided by a sequence of so-called "Strawmen" circulated in advance and intended to provide a concise, modern and expert view of the issues discussed. The third and last of these Strawmen, describing the present state and rationale for ASOF can be found in ASOF Reports & Brochures.
The overall objective of ASOF is:
'To measure and model the variability of fluxes between the Arctic Ocean and the Atlantic Ocean with a view to implementing a longer term system of critical measurements needed to understand the high-latitude ocean's steering role in decadal climate variability.'
Several points are implicit in this statement.
First, in keeping with a number of other current research efforts, it is assumed that the role of the high latitude ocean in decadal climate variability will take effect through its influence on the Atlantic thermohaline circulation (THC or Meridional Overturning Circulation). Most projections of greenhouse gas induced climate change anticipate a weakened MOC in the North Atlantic due to increased freshening and warming in subpolar seas and the supposition is that this climate signal will be transferred to the deep ocean via the two overflows.
Second, it is an underlying assumption of ASOF that in planning to make the first measurements of all the principal oceanic fluxes that connect the Arctic Ocean and North Atlantic through subarctic seas, the point of doing so is 'decadal'. The real objective would not be met until our shorter term research "snapshots" can be set into the context of decadal change.
Third, the ASOF aim suggests there is a point to making these measurements in a coordinated way, so far as possible. For example, a simulated increase in either of the main freshwater outputs that connect the Arctic Ocean with the Atlantic (via the Canadian Arctic Archipelago and along the East Greenland Shelf) seems capable of effecting a slowdown of the MOC and advanced coupled models already indicate that these two main freshwater inputs may have a shared time-dependence. It is appropriate therefore to investigate this finding through simultaneous rather than successive measurement, if at all possible.
The ASOF domain may be described in four ways. The global figure of ongoing and planned process studies in the Atlantic Sector as adopted by the International CLIVAR Project Office (ICPO). It is designed to define the ASOF domain in broad outline in order to set ASOF activities into context of ongoing and planned studies in the Atlantic sector. It correctly defines the primary ASOF focus to be the belt of subarctic seas that connect the Arctic Ocean with the North Atlantic, and makes the point also that many other activities overlap with ASOF in both motivation and location.
Alternately, the figure below defines the ASOF domain in terms of its six main regional tasks, described in the caption. Together, these are intended to meet the ASOF goal of measuring the key ocean exchanges between the Arctic Ocean and subarctic seas, their transformation on passing through the subarctic seas, and their arrival and impact on the overturning circulation of the Northern North Atlantic. Many of these tasks form a successor study to EC-VEINS (1997-2000), but now include attempts to the measure both of the freshwater fluxes through northern seas arguably the most important but least tractable components of exchange.
Implementing this distributed system of measurements will require the funding, access and expertise of agencies and scientists from both North America and Europe, and common funding calls on both continents have been agreed for these sorts of tasks. In order to organise research to meet the available funding on either side of the Atlantic, the ASOF domain and its Steering Group are also organised into "ASOF-East" and "ASOF-West" groupings simply as a practical measure.
The fourth and final subdivision of ASOF, its tasks and domain, is to describe progress towards implementation of the full range of activities region by region that will contribute to each ASOF task. Some of these activities will be directly funded as "ASOF" (for example the measurement of the freshwater flux passing SE Greenland under ASOF-EC). Others will stem from existing nationally or internationally funded efforts that happen to be of central relevance to ASOF aims and objectives. One such is the recent bi-lateral UK-Norway Initiative on Abrupt Climate Change proposed to address this issue by the two Prime Ministers in 1998 which has been developed subsequently into a new thematic programme Rapid Climate Change by the UK Natural Environment Research Council and into the NOClim programme by the Norwegian Research Council. As in the example just given, it is important that we know of all of the main relevant activities in planning the appropriate deployment of effort to meet a given ASOF task.
ASOF will aim to get most of this array into the ocean by the end of 2003 and four factors suggest this is possible. First, certain of the key measurements are already underway. The transport through the Bering Strait has been estimated for decades (Coachman and Aagaard 1988, J. Geophys. Res. 93, 15535-15539) and measured since 1990 (Weingartner et al. 1998, J. Geophys. Res. 103, 7647-7661); the core of the Denmark Strait overflow has been measured with gaps since 1986. Second, in its initial form, the ASOF ISSG is strong in the practical business of maintaining arrays of equipment in these challenging waters. Third, almost all of the techniques needed to make the necessary measurements now seem available or are in prospect. The development at Bedford Institute of Oceanography of the Watson compass for measuring flow directions close to the north magnetic pole, the emergence of a range of cheap profiling CTD systems capable of sub-ice hydrography, the successful trials of sea-glider systems in the past summer giving the prospect of enriching our sparse moored arrays at realistic cost mean that this is possibly the first time that much of ASOF might be achieved.
ASOF Implementation Plan Version 2 (pdf: 36pp, 2MB)