Abstract
Ozone mixing ratios at pressure levels near 2 mbar, derived from Nimbus 4 backscattered ultraviolet (BUV) profile data and averaged within 13 latitude zones between 65°N and 65°S, are studied for the period November 1970 to April 1972. In agreement with previous analyses the largest temporal variations are negatively correlated with changes in zonally averaged equivalent temperature measured simultaneously with the Nimbus 4 selective chopper radiometer. After approximately removing this component using a first‐order photochemical model, residual mixing ratios for latitudes ≲30° contain short‐term variations (periods ≲35 days; rms amplitude ∼1%) that are positively correlated with variations in the solar 10.7‐cm flux and in the 185–190 nm solar ultraviolet flux model of Lean et al. (1982). The correlation coefficients (R = 0.2–0.5; p> 0.95) are larger for each latitude zone when computed versus the ultraviolet flux model than when computed versus the 10.7‐cm flux, suggesting that photochemical responses of upper stratospheric ozone to solar ultraviolet variability at wavelengths near 200 nm are primarily responsible. At latitudes ≳40°, larger‐amplitude wintertime ozone fluctuations associated with planetary‐scale pressure waves become dominant and reduce the computed correlation coefficients to statistically insignificant levels. Linear regression analyses are performed to obtain estimates for the average percent change of ozone at low latitudes and on the considered time scale for given changes in 10.7‐cm flux and in the UV flux model.