Report 27:  A Search for Human Influences on the Thermal Structure of the Atmosphere

Santer, B. D., K. E. Taylor, T. M. L. Wigley, P. D. Jones, D. J. Karoly, J. F. B. Mitchell, A. H. Oort, J. E. Penner, V. Ramaswamy, M. D. Schwarzkopf, R. J. Stouffer and S. Tett
Noivember 1995, 26 pp. 

---

Recent studies have shown that patterns of near-surface temperature change due to combined forcing by carbon dioxide and anthropogenic sulphate aerosols are easier to identify in the observations than signals due to changes in carbon dioxide alone (Santer et al., 1995; Mitchell et al., 1995a). Here we extend this work to the vertical structure of atmospheric temperature changes, and additionally consider the possible effects of stratospheric ozone reduction.

We compare modeled and observed patterns over the lower troposphere to the lower stratosphere (850 to 50 hPa) and over the low- to mid-troposphere (850 to 500 hPa). In both regions there are strong similarities between observed changes and model-predicted signals. Over 850 to 50 hPa similarities are evident both in carbon dioxide-only signals and in signals that incorporate the added effects of sulfate aerosols and stratospheric ozone reduction. These similarities are due largely to a common pattern of stratospheric cooling and tropospheric warming in the observations and model experiments.

Including the effects of stratospheric ozone reduction results in a more realistic height for the transition between stratospheric cooling and tropospheric warming. In the low- to mid-troposphere the observations are in better agreement with the temperature-change patterns due to combined forcing than with the carbon dioxide-only pattern. This is the result of hemispheric-scale temperature-change contrasts that are common to the observations and the combined forcing signal but absent in the carbon dioxide-only case. The levels of model-versus-observed pattern similarity in both atmospheric regions increase over the period 1963 to 1987. If model estimates of natural internal variability are realistic, it is likely that these trends in pattern similarity are partially due to human activities. (pdf file)

UCRL-MI-123395

---