About UNEP UNEP Offices News Centre Publications Events Awards Milestones UNEP Store
GEO Year Book 2004/5  
UNEP Website GEO Home Page
Six steps to abrupt climate change

Theory had already predicted that such changes were possible. In the 1980s, it was suggested that climate warming could add enough freshwater to key places in the oceans to slow or even shut down the thermohaline circulation, leading to reorganization of ocean and atmospheric circulation patterns (Broecker 1987, Broecker and others 1985). Climate model results (Manabe and Stouffer 1988, Rahmstorf 1994) soon lent further support to this theory, and projected substantial cooling in the northern hemisphere, especially in the North Atlantic region, if a shutdown occurred (Figure 3) (Rahmstorf 2002).

Figure 3: Model estimates of air temperature changes resulting from a shutdown of the North Atlantic Conveyor

Source: Rahmstorf 2002

Recent records suggest that the changes predicted by theory and modelling may be actually under way. Measurements of evaporation, precipitation, runoff, ocean salinity, and ocean circulation show these factors changing in ways that may reduce the density of North Atlantic subpolar waters. We may now be observing the early stages of processes that could lead to changes in ocean circulation (Curry and others 1997, Dickson and others 2002, Hansen and others 2001).

The following six steps (Figure 4) lay out one possible sequence of events by which human activities could lead to abrupt climate change.

Figure 4: A possible sequence of events leading to alterations in the North Atlantic thermohaline circulation

Source: Rahmstorf 2002


Figure 5: Changes in Atlantic Ocean salinity distribution from the 1960s to the 1990s

Units are salinity change in parts per thousand (ppt) on log scale as indicated. Black areas are sea bottom features such as the ridges between the Arctic Ocean and North Atlantic near the north end of the transect.

Source: Curry and others 1997

Step 1: Higher carbon dioxide (CO2) emissions increase atmospheric CO2 concentrations.

The burning of fossil fuels (coal, oil and natural gas) and land-use changes have already created a large increase in the concentration of CO2 in the atmosphere. CO2 concentrations have increased by about 35 per cent since the start of the industrial revolution to the current level of 379 parts per million by volume (ppmv) (CDIAC 2004). Concentrations are projected to rise much more if emissions are not sharply reduced (IPCC 2001).

Step 2: This increases global temperatures.

CO2 and other greenhouse gases in the Earth's atmosphere cause an increase in the air temperature near the surface of the Earth. Global average surface air temperature has already risen by 0.6° C over the past 100 years (IPCC 2001). It is projected to rise by another 1.4 to 5.8° C over the next 100 years, according to the range of climate models evaluated by the Intergovernmental Panel on Climate Change (IPCC 2001).

Step 3: Ocean evaporation and surface salinity increase in subtropical latitudes.

The atmospheric warming increases the evaporation of water from the surface of the subtropical oceans, increasing their salinity. A 5-10 per cent increase in evaporation has already been observed in the subtropical Atlantic Ocean over the past 40 years, equivalent to 5-10 cm of surface ocean water each year (Curry and others 1997). Figure 5 shows the resulting increase in surface water salinity in the subtropical Atlantic as calculated and interpolated from direct measurements of salinity. Similar trends in salinity have been observed in the Pacific and Indian Oceans (Wong and others 1999).

Step 4: Precipitation, runoff and glacial melt increase in northern high latitudes, adding excess freshwater to the ocean surface layers in these regions.

The increased moisture evaporated from the subtropical oceans condenses in the atmosphere at higher latitudes, leading to increased precipitation. There has in fact been an increase in precipitation of 6-12 per cent in the northern high latitudes over the last century (IPCC 2001), resulting in increased freshwater runoff from rivers in Russia. The most dramatic increases have occurred in recent decades (Peterson and others 2002) (Figure 6). Increased melting from the Greenland Ice Sheet and other arctic glaciers has also added more freshwater to the Arctic Ocean over the past 40 years (Dyurgerov and Carter 2004). By comparison, the construction of dams and the melting of permafrost have had minor impacts on the long-term pattern of change in river discharge (McClelland and others 2004).

Figure 6: Eurasian river discharge anomaly, and global surface air temperature (SAT) expressed as 10 year running means for 1936–99

Source: Peterson and others 2002

Melting sea ice adds a further source of additional freshwater, because sea ice contains little salt as it rejects most of its salt as it forms. Sea ice extent has declined by 2-3 per cent per decade since 1978 (Comiso and Parkinson 2004). The arctic sea ice is not just shrinking in area but also thinning, leading to predictions that the Arctic Ocean may be free of ice in summer by the end of this century (Yu and others 2004, Laxon and others 2003). These warming-induced increases in precipitation, runoff, glacial melt and sea ice melt could potentially reduce the salinity of surface waters in the Arctic and North Atlantic Oceans.

Step 5: Surface ocean salinity decreases at key locations of deep convection in the North Atlantic.

The Conveyor described above depends on delicately balanced processes. If surface waters in the Greenland, Iceland, Norwegian and Labrador Seas and the subpolar gyre of the North Atlantic are made less salty by an increase in freshwater input due to rising precipitation and runoff, or if temperatures are not sufficiently cold, these waters will not sink as usual. Instead, they will remain on top of the denser saltier waters below, capping them in much the same way as a layer of oil rests above a layer of water. This would stop the initiation of the deep convection that links the surface and bottom portions of the Conveyor.

There is evidence that freshening has been occurring for several decades in the North Atlantic and adjacent seas (Figures 5 and 7). For example, the volume of dense deep water (water of temperature <0.5° C, and of density greater than 1 028 kg/m3) in the Norwegian Sea has been decreasing for the last 50 years. This has led to a decline in the overflow of this deep water (a precursor to North Atlantic Deep Water) via the Faroe Bank Channel into the North Atlantic (Hansen and others 2001) (see Figures 2 and 7).

Figure 7: Declining salinity levels in key areas of the North Atlantic over the last four decades

Source: Modified from Dickson and others 2002

Similarly, the stock of dense deep waters in the Greenland Sea has declined during the period from the 1970s to the 1990s, and a cap of less saline water has accumulated (Curry and others 1997, Curry and Mauritzen in print). The density gradient that drives the overflow across the Denmark Strait Sill has decreased by about 10 per cent, suggesting that the overflow of this second precursor to North Atlantic Deep Water (NADW) may also have declined.

These trends of declining salinity and density in the Nordic Seas are supported by evidence for four decades of salinity decline in deep waters in the North Atlantic and Labrador Sea at additional locations downstream of these overflows (Dickson and others 2002) (Figure 7).

Step 6: There is a slowing or stopping in the ocean circulation that distributes the planet's heat, potentially causing abrupt climate change.

The final step of the process would occur if the sinking of surface water and southward flow of the deep water part of the Conveyor slowed or stopped. If this happened, the warm subtropical waters would not flow northward as they do now.

Direct measurements of a decline in the northward transport of tropical Atlantic Ocean waters have not yet been made, although the multi-decadal slowdown in the overflows of dense deep waters from the Norwegian and Greenland Seas (Hansen and others 2004) suggest that some slowing of the northern-most segment might already be occurring. There is also evidence that some of the freshwater is being carried down and mixed with the deep waters of the western North Atlantic and Labrador Sea (Dickson and others 2002), so while the Conveyor is still operating, it is now carrying more freshwater to depth than in previous decades (Figures 5 and 7).

Earthprint.com Order the Book