IPCC Fourth Assessment Report: Climate Change 2007
Climate Change 2007: Working Group I: The Physical Science Basis

3.5.2 Geopotential Height, Winds and the Jet Stream

Mean changes in geopotential heights resemble in many ways their MSLP counterparts (Hurrell et al., 2004). Linear trends in 700 hPa height during the solstitial seasons, from ERA-40, are shown in Figure 3.24. The 700 hPa level was used as it is the first atmospheric level to lie largely above the East Antarctic Ice Sheet. The NRA and ERA-40 trends agree closely between 1979 and 2001. Over the NH between 1960 and 2000, winter (DJF) and annual means of geopotential height at 850, 500 and 200 hPa decreased over high latitudes and increased over the mid-latitudes, as for MSLP, albeit shifted westward (Lucarini and Russell, 2002). Using NRA, Frauenfeld and Davis (2003) identified a statistically significant expansion of the NH circumpolar vortex at 700, 500 and 300 hPa from 1949 to 1970. But the vortex has contracted significantly at all levels since then (until 2000) and Angell (2006) found a downward trend in the size of the polar vortex from 1963 to 2001, consistent with warming of the vortex core and analysed increases in 850 to 300 hPa thickness temperatures.


Figure 3.24. Linear trends in ERA-40 700 hPa geopotential height from 1979 to 2001 for DJF (top left and bottom right) and JJA (bottom left and top right), for the NH (left) and SH (right). Trends are contoured in 5 gpm per decade and are calculated from seasonal means of daily 1200 UTC fields. Red contours are positive, blue negative and black zero; the grey background indicates 1% statistical significance using a standard least squares F-test and assuming independence between years.

In the NH for 1979 to 2001 during DJF, increases in geopotential height occurred between 30°N and 50°N at many longitudes, notably over the central North Pacific (Figure 3.24). North of 60°N, height changes are consistent with recent occurrences of more neutral phases of the mean polar vortex. Increases in the 700 hPa height outweigh decreases in the northern summer (JJA) during 1979 to 2001. At SH high latitudes, the largest changes are seen in the solstitial seasons (Figure 3.24), with changes of opposite sign in many areas between DJF and JJA. Changes during DJF reflect the increasing strength of the positive phase of the SAM (see Marshall, 2003; Section 3.6.5), with large height decreases over Antarctica and corresponding height increases in the mid-latitudes, through the depth of the troposphere and into the stratosphere. The corresponding enhancement of the near-surface circumpolar westerlies at about 60°S, and associated changes in meridional winds in some sectors, is consistent with a warming trend observed at weather stations over the Antarctic Peninsula and Patagonia (Thompson and Solomon, 2002; see also Sections and 3.6.5). In winter (JJA), there have been height increases over Antarctica since 1979, with a zonal wave 3 to wave 4 pattern of rises and falls in southern mid-latitudes. Trends up to 2001 are relatively strong and statistically significant, with annular modes in both hemispheres strongly positive during the 1990s, although less so in recent years. Hence, geopotential height trends in DJF in the SH through 2004 have weakened in magnitude and significance, but with little change in spatial trend patterns.

Hemispheric teleconnections are strongly influenced by jet streams, which alter waves and storm tracks (Branstator, 2002). Using NRA from 1979 to 1995, Nakamura et al. (2002) found a weakening of the North Pacific winter jet since 1987, allowing efficient vertical coupling of upper-level disturbances with the surface temperature gradients (Nakamura and Sampe, 2002; Nakamura et al., 2004). A trend from the 1970s to the 1990s towards a deeper polar vortex and Iceland Low associated with a positive phase of the NAM in winter (Hurrell, 1995; Thompson et al., 2000; Ostermeier and Wallace, 2003) was accompanied by intensification and poleward displacement of the Atlantic polar frontal jet and associated enhancement of the Atlantic storm track activity (Chang and Fu, 2002; Harnik and Chang, 2003). Analogous trends have also been found in the SH (Gallego et al., 2005).