Knowledge of the sensitivity of climate to external forcing is critical to our ability to predict the effects of anthropogenic greenhouse gases. A recent paper by other authors examined the cumulative response to a series of large volcanic eruptions such as those between 1880 and 1920 using energy-balance modelling and concluded that climate sensitivities of 2.4 degrees for doubled CO2 or more imply volcanic cooling of 0.2 degrees or more continuing well into the 1930s. This long-term cooling is arguably inconsistent with the instrumental temperature record for the first part of this century. If true, this would imply a climate sensitivity significantly lower than the currently accepted "best guess" value of 2.5 degrees.
We have performed similar experiments with a series of large volcanic forcings as input to a simple upwelling-diffusion climate model. Our results do not show the large long-term residual temperature response seen in the previous diffusion-only model work referred to above. Our results for the time period 1880-1950 are reasonably consistent with the temperature record even for climate sensitivity of 4.5 degrees. We discuss the dependence of the model response on model formulation and choice of model parameters, and the implications of these uncertainties for estimates of climate sensitivity