On the quality of the Nimbus 7 LIMS version 6 ozone for studies of the middle atmosphere

The Nimbus 7 Limb Infrared Monitor of the Stratosphere (LIMS) radiance profile dataset of 1978/79 was reconditioned and reprocessed to Version 6 (V6) profiles of temperature and species that are improved significantly over those from Version 5 (V5). The LIMS V6 dataset was archived for public use in 2002. Improvements for its ozone include: (1) a more accurate accounting for instrument and spacecraft motion effects in the radiances, (2) the use of better spectroscopic line parameters for its ozone forward model, (3) retrievals of all its scans, (4) more accurate and compatible temperature versus pressure profiles (or T(p)) that are needed for the registration of the ozone radiances and for the removal of temperature effects from them, and (5) a better accounting for interfering species in the lower stratosphere. The retrieved V6 ozone profiles extend from near cloud top altitudes to about 80 km and from 64S to 84N latitude with better sampling along the orbit than for the V5 dataset. Calculated estimates of the single-profile precision and accuracy are provided; precision estimates based on the data themselves are of order 3% or better from 1 to 30 hPa. Estimates of total systematic error are hard to generalize because the separate sources of error may not all be of the same sign, and they depend somewhat on the atmospheric state. It is estimated that the accuracy of the V6 zonal mean ozone distribution is within ±9% from 50-10 hPa, improving to ±7% in the uppermost stratosphere. Simulation studies show that the LIMS T(p) retrievals are underestimating slightly the small amplitudes of the atmospheric temperature tides, which affect the retrieved day/night ozone differences. There are also small biases in the middle to lower stratosphere for the ascending versus descending node LIMS ozone, due principally to not accounting for the asymmetric weighting of its radiance within the tangent layer. The total accuracy for the LIMS ozone was assessed by comparing its daily zonal mean, daytime distributions against those from the Nimbus 7 SBUV Version 8 (V8) dataset for the same period. The LIMS V6 ozone agrees well with SBUV, except between 2 and 5 hPa where the LIMS ozone is greater. That bias is related to the differing vertical resolutions and forward models for the two experiments. The accuracy for LIMS V6 ozone in the lower stratosphere is improved over that reported for V5, as indicated by a small set of V6 comparisons with ECC ozonesonde profiles. Comparisons of diurnal, photochemical model calculations with the monthly-averaged, upper stratospheric ozone obtained with LIMS V6 indicate only a slight ozone deficit for the model at about 2 hPa. However, that deficit exhibits little to no seasonal variation and is in good agreement with similar model comparisons for a seasonal time series of ozone obtained with ground-based microwave instruments. Because the LIMS V6 ozone has improved accuracy and sampling versus that of V5 for the lower stratosphere it should now be possible to conduct quantitative studies of ozone transport and chemistry for the northern hemisphere, polar winter/spring of 1978/79—a time period when the catalytic loss of ozone due to reactive chlorine should not have been a major factor for the Arctic stratosphere.