Analysis of SAGE II ozone of the middle and upper stratosphere for its response to a decadal-scale forcing

Abstract

Abstract. Stratospheric Aerosol and Gas Experiment (SAGE II) Version 6.2 ozone profiles are analyzed for their decadal-scale responses in the middle and upper stratosphere from September 1991 to August 2005. The profile data are averaged within twelve, 20°-wide latitude bins from 55° S to 55° N and at twelve altitudes from 27.5 to 55.0 km. The separate, 14-yr data time series are analyzed using multiple linear regression (MLR) models that include seasonal, 28 and 21-month, 11-yr sinusoid, and linear trend terms. Proxies are not used for the 28-mo (QBO-like), 11-yr solar uv-flux, or reactive chlorine terms. Instead, the present analysis focuses on the periodic 11-yr terms to see whether they are in-phase with that of a direct, uv-flux forcing or are dominated by some other decadal-scale influence. It is shown that they are in-phase over most of the latitude/altitude domain and that they have max minus min variations between 25° S and 25° N that peak near 4% between 30 and 40 km. Model simulations of the direct effects of uv-flux forcings agree with this finding. The shape of the 11-yr ozone response profile from SAGE II also agrees with that diagnosed for the stratosphere over the same time period from the HALOE data. Ozone in the middle stratosphere of the northern subtropics is perturbed during 1991−1992 following the eruption of Pinatubo, and there are pronounced decadal-scale variations in the ozone of the upper stratosphere for the northern middle latitudes presumably due to dynamical forcings. The 11-yr ozone responses of the southern hemisphere appear to be free of those extra influences. The associated linear trend terms from the SAGE II analyses are slightly negative (−2 to −4%/decade) between 35 and 45 km and nearly constant across latitude. This finding is consistent with the fact that ozone is estimated to have decreased by no more than 1.5% due to the increasing chlorine from mid-1992 to about 2000 but with little change thereafter. It is concluded that a satellite, solar occultation measurement provides both the signal sensitivity and the vertical resolution to record the stratospheric ozone response to the forcing from the solar uv-flux, as well as those due to any other long-term changes.