GCM Simulations of Volcanic Aerosol Forcing. Part I: Climate Changes Induced by Steady-State Perturbations

Abstract

The authors have used the Goddard Institute for Space Studies Climate Model II to simulate the response of the climate system to a spatially and temporally constant forcing by volcanic aerosols having an optical depth of 0.15. The climatic changes produced by long-term volcanic aerosol forcing are obtained by differencing this simulation and one made for the present climate with no volcanic aerosol forcing. These climatic changes are compared with those obtained with the same climate model when the C02 content of the atmosphere was doubled (2?C02) and when the boundary conditions associated with the peak of the last ice age were used (18 K). In all three cases, the absolute magnitude of the change in the globally averaged air temperature at the surface is approximately the same, ?5 K. The simulations imply that a significant cooling of the troposphere and surface can occur at times of closely spaced, multiple, sulfur-rich volcanic explosions that span time scales of decades to centuries, such as occurred at the end of the nineteenth and beginning of the twentieth centuries. The steady-state climate response to volcanic forcing includes a large expansion of sea ice, especially in the Southern Hemisphere; a resultant large increase in surface and planetary albedo at high latitudes; and sizable changes in the annually and zonally averaged air temperature, ?T; ?T at the surface (?Ts) does not sharply increase with increasing latitude, while ?T in the lower stratosphere is positive at low latitudes and negative at high latitudes. In certain ways, the climate response to the three different forcings is similar. Direct radiative forcing accounts for 30% and 25% of the total ?Ts in the volcano and 2?C02 runs, respectively. Changes in atmospheric water vapor act as the most important feedback, and are positive in all three cases. Albedo feedback is a significant, positive feedback at high latitudes in all three simulations, although the land ice feedback is prominent only in the 18 K run. In other ways, the climate response to the three forcings is quite different. The latitudinal profiles of ?Ts for the three runs differ considerably, reflecting significant variations in the latitudinal profiles of the primary radiative forcing. Partially as a result of this difference in the ?Ts profiles, changes in eddy kinetic energy, beat transport by atmospheric eddies, and total atmospheric heat transport are quite different in the three cases. In fact, atmospheric beat transport acts as a positive feedback at high latitudes in the volcano run and as a negative feedback in the other two runs. These results raise questions about the ease with which atmospheric heat transport can be parameterized in a simple way in energy balance climate models.