A Comparison of Southern Hemisphere Circulation Statistics Based on GFDL and Australian Analyses

David J. Karoly National Center for Atmospheric Research, Boulder, CO 80307

Search for other papers by David J. Karoly in
Current site
Google Scholar
PubMed
Close
and
Abraham H. Oort Geophysical Fluid Dynamics Laboratory/NOAA, Princeton University, Princeton, NJ 08542

Search for other papers by Abraham H. Oort in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

Two sets of observed atmospheric circulation statistics for the Southern Hemisphere (SH) are compared. The first set was compiled at the Geophysical Fluid Dynamics laboratory (GFDL) and consists of global objective analyses of circulation statistics accumulated at individual rawinsonde stations for the period May 1963–April 1973. The second set was obtained from daily hemispheric numerical analyses prepared operationally at the World Meteorological Centre, Melbourne, Australia for the period September 1972–August 1982. This study extends the earlier comparison of circulation statistics from station-based and from numerical analysis-based methods by Lau and Oort for the Northern Hemisphere to the Southern Hemisphere.

The domain used for the comparison is a 5° × 5° latitude–longitude grid from 10° to 90°S and seven pressure levels from 1000 to 100 mb. The circulation statistics examined include (i) ten-year averages of the monthly mean fields (measures of the mean circulation), (ii) ten-year averages of the standard deviations and covariances of daily values (measures of the daily transient eddy variability) and (iii) year-to-year standard deviations of the monthly mean fields (measures of the interannual variability). The statistics are presented using horizontal maps on pressure surfaces and latitude–pressure sections of zonal averages.

The two sets of circulation statistics were derived using very different analysis methods and they apply for different time periods. The similarities and differences between the statistics from the two datasets indicate the reliability of the statistics and can be used to define a better composite set of circulation statistics for the SH.

The relatively large differences in the statistics can generally be attributed to the sparse conventional observation network in the SH, particularly over the large ocean regions, and deficiencies in the analysis methods. The two sets agree reasonably well from 850 to 500 mb over the land masses, where the observation network is less sparse. In the upper troposphere, the magnitudes of the daily transient eddy statistics from the Australian dataset are smaller due to the analysis method and the inclusion of satellite data. Over the data-sparse regions, the use of the zonal average as the first guess for the GFDL dataset has led to reduced spatial variability, smoother fields and underestimation of extreme values.

Abstract

Two sets of observed atmospheric circulation statistics for the Southern Hemisphere (SH) are compared. The first set was compiled at the Geophysical Fluid Dynamics laboratory (GFDL) and consists of global objective analyses of circulation statistics accumulated at individual rawinsonde stations for the period May 1963–April 1973. The second set was obtained from daily hemispheric numerical analyses prepared operationally at the World Meteorological Centre, Melbourne, Australia for the period September 1972–August 1982. This study extends the earlier comparison of circulation statistics from station-based and from numerical analysis-based methods by Lau and Oort for the Northern Hemisphere to the Southern Hemisphere.

The domain used for the comparison is a 5° × 5° latitude–longitude grid from 10° to 90°S and seven pressure levels from 1000 to 100 mb. The circulation statistics examined include (i) ten-year averages of the monthly mean fields (measures of the mean circulation), (ii) ten-year averages of the standard deviations and covariances of daily values (measures of the daily transient eddy variability) and (iii) year-to-year standard deviations of the monthly mean fields (measures of the interannual variability). The statistics are presented using horizontal maps on pressure surfaces and latitude–pressure sections of zonal averages.

The two sets of circulation statistics were derived using very different analysis methods and they apply for different time periods. The similarities and differences between the statistics from the two datasets indicate the reliability of the statistics and can be used to define a better composite set of circulation statistics for the SH.

The relatively large differences in the statistics can generally be attributed to the sparse conventional observation network in the SH, particularly over the large ocean regions, and deficiencies in the analysis methods. The two sets agree reasonably well from 850 to 500 mb over the land masses, where the observation network is less sparse. In the upper troposphere, the magnitudes of the daily transient eddy statistics from the Australian dataset are smaller due to the analysis method and the inclusion of satellite data. Over the data-sparse regions, the use of the zonal average as the first guess for the GFDL dataset has led to reduced spatial variability, smoother fields and underestimation of extreme values.

Save