Air Pollution as a Climate Forcing: A Workshop
Day 1 Presentations
Climatic Implications of Changes in Tropospheric Ozone
Loretta J. Mickley*, Daniel J. Jacob*, and David Rind+
* Harvard University, Cambridge, MA, U.S.A.
+ NASA Goddard Institute for Space Studies, New York, NY, U.S.A.
You may download a MS PowerPoint version (6.2 MB) of this presentation.
Radiative forcing is a yardstick commonly used to assess the relative importance to climate change of changes in the concentrations of greenhouse gases and aerosols. In particular, recent model calculations of the global mean radiative forcing from tropospheric ozone since preindustrial times yield values in a relatively narrow range, from 0.3 to 0.5 W/m2, or about one-fourth to one-third the forcing due to CO2. These results imply that (1) the radiative forcing due to added tropospheric ozone is relatively well-known, and that (2) the response of surface temperature to added ozone is also about one-fourth to one-third that of CO2.
However, regarding the first implication, the forcing calculations use preindustrial ozone fields that overestimate observations available from the turn of the nineteenth century. We show that a global three-dimensional model of tropospheric chemistry with reduced nitrogen oxide emissions from lightning (1-2 Tg N/y) and soils (2 Tg N/y), and increased emissions of biogenic hydrocarbons can better reproduce the nineteenth century observations. The resulting global mean radiative forcing from tropospheric ozone since preindustrial times is 0.72-0.80 W/m2, amounting to about half of the estimated CO2 forcing. Our results indicate that the uncertainty in computing radiative forcing from tropospheric ozone since preindustrial times is larger than is usually acknowledged.
Regarding the second implication, we show using a version of the same model that the sensitivity of globally averaged temperature at the surface to a given forcing from tropospheric ozone is about three-fourths the sensitivity to the same forcing from carbon dioxide. We performed a set of equilibrium climate simulations, using the "qflux" version of the GISS GCM, which allows the response of sea surface temperature. The following simulations were performed: (1) a control, (2) a control with CO2 reduced by 25 ppm, (3) a simulation in which monthly averaged fields of previously calculated, present-day ozone were permitted to influence climate, (4) a corresponding simulation using fields of preindustrial ozone, and finally (5) a simulation using calculated preindustrial ozone increased by 18 ppb everywhere. We decreased CO2 by 25 ppm for run 2 since an increase of 25 ppm CO2 yields the same globally averaged forcing as the change in tropospheric ozone since preindustrial times. For run 5, the uniform increase of ozone by 18 ppb corresponds to the globally averaged increase in ozone mixing ratio from preindustrial times to the present-day.
We calculate an equilibrium temperature response of about +0.4° C for the 25-ppm change in CO2, compared to +0.3° C for the change in tropospheric ozone from preindustrial times to the present-day. In the lower stratosphere, however, the increased ozone leads to a cooling of about 0.2° C, compared to a slight warming (0.1° C) with increased CO2. Both the "realistic" present-day ozone fields (run 3) and the fields with tropospheric ozone increased uniformly (run 5) lead to greater warming at the surface in the Northern Hemisphere relative to the Southern Hemisphere, but the interhemispheric difference is about double (nearly 0.2° C) with the realistic ozone due to greater industrial activity and ozone concentrations in the Northern Hemisphere.
Our work suggests that the apparently greater sensitivity of surface temperature to a CO2 forcing than to a forcing from tropospheric ozone has two causes. First, for the same globally averaged longwave forcing, CO2 forcings are stronger than those of ozone at the poles, where climate is most sensitive, than over the tropics due to the interference of low-latitude water vapor and high, cold clouds. Second, as the poles warm in response to either increased CO2 or ozone in the model, the ice and snow albedo decreases, which means that the shortwave forcing of ozone, which is greatest over reflective surfaces, diminishes. In effect, the well-known positive albedo feedback at the poles is dampened in the case of increasing ozone, since the shortwave forcing of ozone depends on the albedo.