A three-dimensional simulation of the evolution of the middle latitude winter ozone in the middle stratosphere

Observations of O-3, CH4, and N2O from the Halogen Occultation Experiment (HALOE), the Cryogenic Limb Array Etalon Spectrometer (CLAES), and the Microwave Limb Sounder on the Upper Atmosphere Research Satellite are compared with fields calculated using a three-dimensional global chemistry and transport model which uses winds from a data assimilation system. Model and observations are found to be in generally good agreement for ozone at middle latitudes between 50 and 3 hPa. Much of the horizontal structure in the observed long-lived tracers N2O and CH4 is seen in the model, but the model profile shapes differ systematically from those seen by CLAES (N2O and CH4) and HALOE (CH4). To understand the influence of transport on these comparisons, attention is focused on January and February 1992, when a series of minor warmings impact the ozone and tracer distributions. The model O-3 at 6.8 hPa, which agrees well with observations, is shown to be influenced by transport from the tropics toward middle and high latitudes, which increases O-3, and photochemical processes, through which O-3 is destroyed. However, the model N2O and CH4 mixing ratios at middle latitudes at 6.8 hPa are much greater than observations. Good comparisons between modeled and observed O-3 and poor comparisons between modeled and observed N2O and CH4 are explained by differences in the horizontal transport from the subtropics. Here photochemical processes control the O-3 mixing ratio, and the horizontal winds transport air with realistic ozone mixing ratios to the middle latitudes. For the longer-lived gases, the model mixing ratios in the subtropics far exceed observations, and the horizontal winds transport air with higher than observed mixing ratios to middle latitudes, leading to the poor comparisons. This analysis shows that the photochemical-and transport processes that affect O-3 at middle latitudes during winter are well represented in the model. However, for long-lived tracers, errors in the model subtropics are communicated to middle latitudes by the horizontal winds.