ABRUPT CLIMATE CHANGE & THE ATLANTIC OCEAN: MONITORING & PREDICTING THE ATLANTIC HEAT CONVEYOR

Fig. 4

Fig.3

Fig.2

Fig.1

by Dr. Craig Wallace
Dr. Craig Wallace is the NERC RAPID program climate change knowledge transfer manager and he is based at the National Oceanography Centre (NOC), Southampton

One of the more media-appealing scenarios related to global climate change is the notion that changes in the temperature and saltiness of the Atlantic Ocean will slow down, or even halt, the Atlantic ‘Gulf Stream’ system. The pitfalls of overexposing this threat via the modern media are all too well known – take for example the hyperbolical film The Day After Tomorrow, which did as much to confuse it’s audience as it did to entertain. What, then, are the key issues surrounding this intriguing aspect of the climate system? Is there a threat at all? Is the threat big? And, fundamentally, what can we do to learn more? To answer these questions the UK Natural Environment Research Council launched a dedicated thematic research programme, RAPID, in 2003 with a mission to ‘increase our ability to quantify the probability and magnitude of future abrupt climate change’. Although strongly focussed on the threat posed by re-arrangement of the Atlantic Ocean circulation, the programme is also investigating the potential threat from other features of the climate system, such as the El-Nino, Southern Oscillation phenomenon. The programme funds approximately 40 individual research projects, primarily within the UK, but also maintains strong links with scientists from the US, Holland and Germany. Much of RAPID’s £20 million budget has been used to deploy and maintain a network of moored sensors across the Atlantic Ocean [Fig. 1] in an effort to develop a system capable of monitoring the strength of the so-called meridional overturning circulation, of which the Gulf Stream is a part. Initial results are encouraging suggesting that we can, in fact, observe the strength of the circulation, using this distribution of equipment, to a high degree of accuracy.

Fig.1: Locations of RAPID Atlantic moorings designed to monitor the strength and variability of the ‘meridional overturning circulation’. Equipment was deployed in 2004 and data is recovered every 6 months.

With funding in place until 2008/2009 RAPID moorings will deliver the first, continuous multi-year time series of circulation strength, vital if we are to understand what the year-to-year variability of the system is. Knowledge of this variability is crucial if we ever hope to detect, with certainty, any long-term trend. For example, a comparison of five sets of ship-based measurements, by another RAPID project, suggests that the circulation has slowed by as much as 30% between 1957 to 2004. But without knowledge of how the system varies naturally from year-to-year it is impossible to say whether this is a bona-fide signal, or whether the ships have just so happened to observe 5 measurements which give this limited, and wrong, impression of the system [Fig 2].

Fig.2: The dangers of implying long-term trends over a time series whose variability you know little about. The five observations (red), in a and b, both suggest a downward trend: correct for a, but misleading in b, where there is high variability and actually a rising trend.

A significant number of RAPID projects are also working to refine the ocean models used to predict future changes to the Gulf Stream – introducing new features, Arctic polynyas (semipermanent areas of open water in sea ice)for example, or by identifying the physical reasons why the Gulf Stream behaves so differently in each model during global warming experiments [Fig. 3]. Working to reduce this uncertainty is a major goal for RAPID and a suite of experiments using an ensemble of the UK’s leading climate models is underway. Results from this experiment can also be used to further investigate the actual climatic impacts of both a slowing and shutdown of the circulation.

Fig. 3: Different climate model projections for future ‘meridional overturning circulation’ strength (change in flow strength from pre-industrial climate) under global warming, up to 2100. The spread of future strength is large, but most models agree a weakening is likely. A comparison of the UK’s leading climate models is underway in RAPID. Source: IPCC, 2001.

The task of predicting the future behaviour of the Gulf Stream system, or any other part of the climate system, is further complicated by the fact that repeating identical global warming experiments with the same climate model results in a slightly different climate state. This is due to the model’s ability to replicate some of the inherent ‘randomness’ of the real world climate system and has resulted in the current preference in climate science to present results in the form of ‘probability distributions’ [Fig. 4]. RAPID scientists are developing powerful statistical techniques to enable the production of forecasts of future Gulf Stream strength in this style, but without having to re-run climate models repeatedly – an expensive exercise in both time and cost.

Fig. 4: An example ‘probability density function’, a favoured way to present the probability of future climate change. The horizontal axis represents a range of future values of the variable of interest (e.g. global mean temperature) and the y axis, the probability of each value happening.

Despite the complexities and challenges related to this topic, progress within the field is, nonetheless, steady. The recent Intergovernmental Panel on Climate Change Fourth Assessment Report, to which RAPID science contributed, was able to say with 90% confidence that a slow down in Gulf Stream strength will occur this century, and that the range of predicted reduction is between 0 and 50% - not enough to lead to local cooling. Conversely, the prospect of a shutdown this century remains unlikely. Of course, whilst climate models remain our most useful tool at present with which to plan for the future, our trust in them can only be consolidated by knowing that they realistically resemble the real world. It is here, principally, by careful comparison and utliilisation of our data that we hope RAPID can elevate our understanding of this important topic to the next level.

(Posdata del editor del blog: Es evidente que las conclusiones del profesor Wallace son pocas y de escasa consistencia, al dejar pendiente de múltiples condiciones su criterio final. En todo caso habría que ver si las tomas de mediciones, con el deshielo que se esta produciendo en Groenlandia y el Ártico durante este verano del 2009 nos dará suficiente información sobre el cambio de la salinidad y la temperatura suficientes para que las previsiones, siempre muy atrasadas de información, de Panel para el Cambio Climatico de las Naciones Unidas-IPCC- se queden obsoletas una vez más).