3.5.3 Storm Tracks
A number of recent studies suggest that cyclone activity over both hemispheres has changed over the second half of the 20th century. General features include a poleward shift in storm track location, increased storm intensity, but a decrease in total storm numbers (e.g., Simmonds and Keay, 2000; Gulev et al., 2001; McCabe et al., 2001). In the NH, McCabe et al. (2001) found that there has been a significant decrease in mid-latitude cyclone activity and an increase in high-latitude cyclone frequency, suggesting a poleward shift of the storm track, with storm intensity increasing over the North Pacific and North Atlantic. In particular, Wang et al. (2006a) found that the North Atlantic storm track has shifted about 180 km northward in winter (JFM) during the past half century. The above findings are corroborated by Paciorek et al. (2002), Simmonds and Keay (2002) and Zhang et al. (2004b).
Several results suggest that cyclone activity in the NH mid-latitudes has increased during the past 40 years. Increases in storm track activity have been found in eddy statistics, based on NRA data. North Pacific storm track activity, identified as poleward eddy heat transport at 850 hPa, was significantly stronger during the late 1980s and early 1990s than during the early 1980s (Nakamura et al., 2002). A striking signal of decadal variability in Pacific storm track activity was its midwinter enhancement since 1987, despite a concurrent weakening of the Pacific jet, concomitant with the sudden weakening of the Siberian High (Nakamura et al., 2002; Chang, 2003). Significant increasing trends over both the Pacific and Atlantic are found in eddy meridional velocity variance at 300 hPa and other statistics (Chang and Fu, 2002; Paciorek et al., 2002). Since 1980, there was an increase in the amount of eddy kinetic energy in the NH due to an increase in the efficiency in the conversion from potential to kinetic energy (Hu et al., 2004). Graham and Diaz (2001) also found an increase in MSLP variance over the Pacific.
There are, however, significant uncertainties with such analyses, with some studies (Bromirski et al., 2003; Chang and Fu, 2003) suggesting that storm track activity during the last part of the 20th century may not be more intense than the activity prior to the 1950s. Eddy meridional velocity variance at 300 hPa in the NRA appears to be biased low prior to the mid-1970s, especially over east Asia and the western USA (Harnik and Chang, 2003). Hence, the increases in eddy variance in the NRA reanalysis data are nearly twice as large as that computed from rawinsonde observations. Better agreement is found for the Atlantic storm track exit region over Europe. Major differences between radiosonde and NRA temperature variance at 500 hPa over Asia (Iskenderian and Rosen, 2000; Paciorek et al., 2002) also cast doubts on the magnitude of the increase in storm track activity, especially over the Pacific.
Station pressure data over the Atlantic-European sector (where records are long and consistent) show a decline of storminess from high levels during the late 19th century to a minimum around 1960 and then a quite rapid increase to a maximum around 1990, followed again by a slight decline (Alexandersson et al., 2000; Bärring and von Storch, 2004; see also Section 220.127.116.11). However, changes in storm tracks are expected to be complex and depend on patterns of variability, and in practice, the noise present in the observations makes the detection of long-term changes in extratropical storm activity difficult. A more relevant approach seems to be the analysis of regional storminess in relation to spatial shifts and strength changes in teleconnection patterns (see Section 3.6).
Significant decreases in cyclone numbers, and increases in mean cyclone radius and depth over the southern extratropics over the last two or three decades (Simmonds and Keay, 2000; Keable et al., 2002; Simmonds, 2003; Simmonds et al., 2003) have been associated with the observed trend in the SAM. Such changes, derived from NRA data, have been related to reductions in mid-latitude winter rainfall (e.g., the drying trend observed in south-western Australia (Karoly, 2003) and to a circumpolar signal of increased precipitation off the coast of Antarctica (Cai et al., 2003). However, there are significant differences between ERA-40 and NRA in the SH: higher strong-cyclone activity and less weak-cyclone activity over all oceanic areas south of 40°S in all seasons, and stronger cyclone activity over the subtropics in the warm season in ERA-40, especially in the early decades (Wang et al., 2006a).