Accounting for the Effects of Volcanoes and ENSO in Comparisons of Modeled and Observed Temperature Trends

B. D. Santer, C. Doutriaux, J. S. Boyle, K. E. Taylor

Program for Climate Model Diagnosis and Intercomparison, Lawrence Livermore National Laboratory, Livermore, CA 94550, U.S.A.

T. M. L. Wigley, G. A. Meehl

National Center for Atmospheric Research, Boulder, Colorado

J. E. Hansen

NASA/Goddard Institute for Space Studies, New York

P. D. Jones

Climatic Research Unit, University of East Anglia, Norwich, U.K.

E. Roeckner

Max-Planck Institute for Meteorology, D-20146 Hamburg, Germany

S. Sengupta

Engineering Directorate, Lawrence Livermore National Laboratory, Livermore, CA 94550, U.S.A.


Abstract

Several previous studies have attempted to remove the effects of explosive volcanic eruptions and El Nino/Southern Oscillation (ENSO) variability from thime series of globally averaged surface and tropospheric temperatures. Such work has largely ignored the possibility of non-zero correlation between volcanic signals and ENSO, a form of collinearity. Here we account for this collinearity using an iterative procedure. We remove the estimated volcano and ENSO signals from the raw global mean temperature data, and then calculate trends in the residuals.

For observational temperature data, sensitivities to the choice of ENSO index and to uncertainties in key volcanic parameters lead to ranges of 'residual' trends at the surface (0.214 to 0.245oC/decade), in the lower troposphere (2LT; 0.068 to 0.186oC/decade), and in the surface-minus-lower troposphere trend differential (0.060 to 0.144oC/decade). Residual surface and lower tropospheric trends are invariably larger than trends in the raw obervational data. The residual surface-minus-2LT trend difference is generally smaller than in the raw data. These results suggest that volcanoes and ENSO significantly depressed both globally-averaged surface and tropospheric temperatures over 1979-1999, and exerted a larger cooling influence on the troposphere than on the surface.

ENSO and volcanic forcing effects can hamper reliable assessment of the true correspondence between modeled and observed trends. In the second part of our study, we use our iterative approach to remove these effects from model data. After removal, the residual modeled and observed global-mean surface temperature trends are in good agreement. For the lower troposphere, the statistical significance of differences between modeled and observed residual trends depends on the experiment considered, the choice of ENSO index, and the volcanic signal decay time. Residual surface-minus-2LT trend differences are significantly smaller than in observations in 11 out of 12 cases considered.

Finally, we consider results from recent model experiments with relatively complete estimates of natural and anthropogenic forcing. As for residual trends, we find close agreement between raw observed and simulated surface temperature trends. In 25 out of 30 cases, simulated lower tropospheric trends are not statistically different from observations. While simulated surface-minus-2LT trend differentials are significantly smaller than observed, they are positive (larger warming aloft than at the surface) in 5 of 30 cases, behavior qualitatively similar to observations.


Back to Publications List

To the Climate Analysis Section Home Page


Hongjun Zhang: zhangho@ucar.edu