Have the Southern Hemisphere storm tracks been strengthening after 1979?

Based on both the NCEP-NCAR and ERA40 reanalyses data, storm track activity in the Southern Hemisphere (SH), as measured by eddy variance and covariance statistics, has strengthened significantly since the late 1950s. Several previous studies have suggested that this long term trend may be significantly biased due to the introduction of satellite data in the 1970s that significantly impacted the data assimilation systems, such that climate trends in reanalysis data that straddle the pre- and post- satellite eras may not be reliable. Nevertheless, embedded within this long term trend, SH storm track activity computed from the NCEP-NCAR reanalysis data shows a slight poleward shift as well as significant increase (by about 20%) even after 1979, during the satellite era. However, a similar increasing trend is not found from ERA40 data, with the storm tracks in the SH in ERA40 displaying a slight poleward shift but no significant increasing trend.

Based on IPCC AR4 model simulations, storm track activity, especially over the SH, is expected to shift poleward and intensify slightly in the upper troposphere under various global warming scenarios. However, none of the AR4 simulations that we have examined show significant SH storm track increase between 1979 and 2000, even though some simulations do suggest a slight poleward shift during this period. Thus assessing whether there has been a significant trend in storm track activity between 1979 and 2000 is important to see whether AR4 climate simulations are able to reproduce past climate variability and trends, as well as to assess whether the increasing trend in storm track activity predicted by AR4 models may represent an underestimation.

Currently, we are analyzing diagnostic fields from the reanalysis datasets to examine whether the increasing trend (or its absence) is consistent with various climate diagnostics. Preliminary results based on examination of energetics show that while eddy kinetic energy (EKE) shows an increasing trend in the NCEP-NCAR reanalysis data, the EKE source terms (baroclinic and barotropic conversion) do not show a consistent trend, suggesting that the NCEP-NCAR reanalysis trend may be internally inconsistent. Other diagnostics, such as reanalysis generated precipitation, clouds, and radiation fields, are being examined and compared with available observational data. By the time of the conference we should have a more definitive assessment of whether the trend found in the NCEP-NCAR reanalysis is realistic or not.