Climate Dynamics

, Volume 23, Issue 2, pp 207–214

Quantifying the water vapour feedback associated with post-Pinatubo global cooling


    • NOAA Aeronomy Laboratory
    • Department of MeteorologyUniversity of Reading
  • M. Collins
    • Centre for Global Atmospheric ModellingDepartment of Meteorology, University of Reading

DOI: 10.1007/s00382-004-0431-z

Cite this article as:
de F. Forster, P.M. & Collins, M. Climate Dynamics (2004) 23: 207. doi:10.1007/s00382-004-0431-z


There is an ongoing important debate about the role of water vapour in climate change. Predictions of future climate change depend strongly on the magnitude of the water vapour feedback and until now models have almost exclusively been relied upon to quantify this feedback. In this work we employ observations of water vapour changes, together with detailed radiative calculations to estimate the water vapour feedback for the case of the Mt. Pinatubo eruption. We then compare our observed estimate with that calculated from a relatively large ensemble of simulations from a complex coupled climate model. We calculate an observed water vapour feedback parameter of –1.6 Wm–2 K–1, with uncertainty placing the feedback parameter between –0.9 to –2.5 Wm–2 K–1. The uncertain is principally from natural climate variations that contaminate the volcanic cooling. The observed estimates are consistent with that found in the climate model, with the ensemble average model feedback parameter being –2.0 Wm–2 K–1, with a 5–95% range of –0.4 to –3.6 Wm–2 K–1 (as in the case of the observations, the spread is due to an inability to separate the forced response from natural variability). However, in both the upper troposphere and Southern Hemisphere the observed model water vapour response differs markedly from the observations. The observed range represents a 40%–400% increase in the magnitude of surface temperature change when compared to a fixed water vapour response and is in good agreement with values found in other studies. Variability, both in the observed value and in the climate model’s feedback parameter, between different ensemble members, suggests that the long-term water vapour feedback associated with global climate change could still be a factor of 2 or 3 different than the mean observed value found here and the model water vapour feedback could be quite different from this value; although a small water vapour feedback appears unlikely. We also discuss where in the atmosphere water vapour changes have their largest effect on surface climate.

Copyright information

© Springer-Verlag  2004