Closing comments

What is the future of the Arctic? The ACIA projections point to a warmer and wetter regime with a fresher Arctic Ocean and less sea ice. One of the scenarios indicates a complete loss of September ice by around 2070. A review of the observational records for the past several decades reveals that temperatures have been rising and sea ice has been declining. There is evidence of increased precipitation and river discharge. But there are key differences in comparison to the model projections. The models project the strongest warming over the Arctic Ocean during autumn and winter. Direct observations point to the strongest warming over land areas during winter and spring. Satellite records point to warming in all seasons, but with cooling in some areas. Trends, however, are quite sensitive to the exact period examined. While there has been some warming over the Arctic Ocean, which has been associated with sea ice losses, it is clear that the spatial pattern of comparatively larger changes projected for this region has yet to be realized. Serreze and Francis (2005) argue that when recent SAT trends are compared to ACIA model projections for a near-future time slice (2010-29) there are no fundamental disparities. They suggest the Arctic Ocean can be viewed as in a

CGC - CSM - ECH GFD HAD

2000 2010 20202030 2040 2050 2060 2070 2080 2090 2000 2010 2020 2030 20402050 2060 2070 2030 2090

Figure 11.24 Time series of Northern Hemisphere sea ice extent (at least 15% ice cover), March and September, for the period 2000 through 2100 from the five AOGCMs participating in the Arctic Climate Impact Assessment. The top panels represent the raw model output. The bottom panels adjust the output based on the model bias relative to observations (see text) (courtesy of J. Walsh and W. Chapman, University of Illinois, Urbana-Champaign, IL).

2000 2010 20202030 2040 2050 2060 2070 2080 2090 2000 2010 2020 2030 20402050 2060 2070 2030 2090

Figure 11.24 Time series of Northern Hemisphere sea ice extent (at least 15% ice cover), March and September, for the period 2000 through 2100 from the five AOGCMs participating in the Arctic Climate Impact Assessment. The top panels represent the raw model output. The bottom panels adjust the output based on the model bias relative to observations (see text) (courtesy of J. Walsh and W. Chapman, University of Illinois, Urbana-Champaign, IL).

state of preconditioning, associated with initial retreat and thinning of its sea ice cover, and that a threshold will be reached in the next few decades after which we will see strong autumn and winter warmings.

This brings us back to the problem of low-frequency variability. Within the available instrumental record, the Arctic has exhibited considerable variability on decadal and multidecadal time scales. While a great deal of emphasis has been placed on the terrestrial warming since about 1970, it must be remembered that the rise in surface air temperatures from about 1920 to 1940 was just as large, which was then followed by a period of cooling. As has been frequently pointed out (e.g., Serreze et al., 2000; Polyakov and Johnson, 2000; Polyakov et al., 2002; Semenov and Bengtsson, 2003) such low-frequency variability in the Arctic can make it very difficult to separate natural fluctuations in Arctic climate from those due to trace gas loading.

Delworth and Knutson (2000) attempted to explain the mechanisms of the earlier twentieth-century warming with a coupled ocean-atmosphere climate model. They included observed, time-varying concentrations of greenhouse gases and sulfate aerosols in six realizations of the past 135 years. These runs were compared to those from a 1000-year control simulation with a constant atmospheric composition. The earlier twentieth-century warming event was reproduced by only one of the six realizations, indicating that natural climate variability was the major contributor. But, as warming events of this magnitude were not seen in the control run, they concluded that greenhouse gases and sulfate aerosols may have played a role. This event appears to have been associated with an increase in poleward oceanic heat transport in the North Atlantic, with consequent reductions in ice extent. The observational analysis of Semenov and Bengtsson (2003) lends further support to a linkage between ocean heat transport and sea ice reductions, as does the subsequent modeling study of Bengtsson et al. (2004). The winter NAO was in a general, albeit not remarkably, positive phase during this period (Figure 11.9).

An NAO/AO link with many of the recently observed Arctic changes is well established. Could greenhouse gas loading, ozone losses or other anthropogenic influences work to favor the positive mode of the NAO/AO? While there is evidence to support this view, the issue is far from resolved. Regardless, low-frequency variability internal to the coupled atmosphere-ocean system or from natural external forcing will always be part of the mix. Regarding the latter, Ogi et al. (2003) find a link between the NAO and the 11-year solar cycle, which presumably involves stratospheric responses to altered ultraviolet radiation. Further support for a solar link comes from the modeling study of Rind et al. (2004) for climate cooling during the Maunder Minimum (1645-1715). During this period of low solar activity (few sunspots), most of their experiments produce a relative negative phase of the NAO/AO. However, such solar variability/climate links remain to be unequivocally established. Interestingly, recent years have seen a shift of the NAO/AO toward a more neutral state. Is it entering a new negative regime? If so, will this result in temporary cooling, or will the effects be superimposed on a more general greenhouse-gas warming trend?

The fundamental physics behind the climate model projections is basically sound. To our minds, it is not really a question of whether the Arctic will change in response to human activities, but rather when and by how much. While current-day models tend to project rather large changes, many processes are still not well represented. If, for example, there is a significant weakening of the THC, temperature rises, at least regionally, might be dampened. This effect could be greater than we think - at least one model simulation, which includes coupling with the Greenland Ice Sheet, suggests a strong and abrupt weakening of the THC at the end of the twenty-first century triggered by enhanced freshwater input arising from partial melt of the Greenland Ice Sheet, with marked cooling over eastern Greenland and the northern North Atlantic (Fichefet et al., 2003).

Better understanding of the future course of the Arctic climate system requires improved models, but equally important, a desire to break down traditional disciplinary barriers and develop a more system oriented approach. The Arctic is the home of myriad climate interactions and feedbacks involving the atmosphere, land, ocean and cryosphere, which are in turn tightly coupled to the global climate system. But it seems that the more we understand how the Arctic functions, the more questions we raise. Herein lies our continued fascination with the Arctic climate system.

0 0

Post a comment