75
mean surface air temperature (Hurrell 1996), and they account for virtually
all the accelerated wintertime warming over the high latitude continents
reported in the IPCC assessments (Wallace et al. 1995a,b). Hence, it
seems unlikely that this 'accelerated warming' can continue much longer.
In fact, if the interdecadal component of either the NAO or ENSO undergoes a polarity reversal during the next decade, winter temperatures
over the high latitude continents could actually fall, even in the presence
of a modest upward trend in global-mean temperature. In view of the
much greater sensitivity of cold season temperatures to the vicissitudes of
the coupled atmosphere-ocean climate system, warm season data may be
a more reliable indicator of current, anthropogenic ally induced trends in
hemispheric-mean temperature.
Some readers may question our premise that interdecadal-scale variations in the atmospheric circulation should be listed among the possible
causes of the recent rise in hemispheric-mean surface air temperature. The
inverse view (i.e., that the recent rise in hemispheric-mean temperature
is responsible for the circulation anomalies that have prevailed during recent years) is certainly not without adherents. For example, Namias (1980)
computed correlation patterns analogous to those in our Fig. 13 for the expressed purpose of demonstrating possible regional effects associated with
global warming. In defense of our interpretation, we have offered a physically plausible argument as to why the presence of warm anomalies over
the continents and cold anomalies over the oceans should favour anomalously warm hemispheric-mean surface air temperature, whereas it is not
at all clear how a rise in hemispheric-mean temperature of a mere few
tenths of a degree K could force the large, regionally specific atmospheric
circulation changes that have been observed during the past two decades.
Furthermore, although some records were broken, we are not convinced
that the amplitudes of either the Southern Oscillation or the North Atlantic Oscillation during the 1980s and early 90s were large enough to to
be considered beyond the range of natural variability. Time will tell which
of the two interpretations is more valid.
10 Acknowledgments
Figure 2 was inspired by an workshop presentation by Henry Diaz and
Fig. 3 by published results of Bradley et al. (1987). Figures 1, 5, 6, 8,
11 and 12 were kindly provided by Xinhua Chen; Figs. 2-4, 13-17, 24,
mean surface air temperature (Hurrell 1996), and they account for virtually
all the accelerated wintertime warming over the high latitude continents
reported in the IPCC assessments (Wallace et al. 1995a,b). Hence, it
seems unlikely that this 'accelerated warming' can continue much longer.
In fact, if the interdecadal component of either the NAO or ENSO undergoes a polarity reversal during the next decade, winter temperatures
over the high latitude continents could actually fall, even in the presence
of a modest upward trend in global-mean temperature. In view of the
much greater sensitivity of cold season temperatures to the vicissitudes of
the coupled atmosphere-ocean climate system, warm season data may be
a more reliable indicator of current, anthropogenic ally induced trends in
hemispheric-mean temperature.
Some readers may question our premise that interdecadal-scale variations in the atmospheric circulation should be listed among the possible
causes of the recent rise in hemispheric-mean surface air temperature. The
inverse view (i.e., that the recent rise in hemispheric-mean temperature
is responsible for the circulation anomalies that have prevailed during recent years) is certainly not without adherents. For example, Namias (1980)
computed correlation patterns analogous to those in our Fig. 13 for the expressed purpose of demonstrating possible regional effects associated with
global warming. In defense of our interpretation, we have offered a physically plausible argument as to why the presence of warm anomalies over
the continents and cold anomalies over the oceans should favour anomalously warm hemispheric-mean surface air temperature, whereas it is not
at all clear how a rise in hemispheric-mean temperature of a mere few
tenths of a degree K could force the large, regionally specific atmospheric
circulation changes that have been observed during the past two decades.
Furthermore, although some records were broken, we are not convinced
that the amplitudes of either the Southern Oscillation or the North Atlantic Oscillation during the 1980s and early 90s were large enough to to
be considered beyond the range of natural variability. Time will tell which
of the two interpretations is more valid.
10 Acknowledgments
Figure 2 was inspired by an workshop presentation by Henry Diaz and
Fig. 3 by published results of Bradley et al. (1987). Figures 1, 5, 6, 8,
11 and 12 were kindly provided by Xinhua Chen; Figs. 2-4, 13-17, 24,