The radiative infrared cooling of CO₂ in the middle atmosphere is crucial for computing its thermal structure. It requires one however to include non-local thermodynamic equilibrium processes which are computationally very expensive, which cannot be afforded by climate models. In this work, we present an updated, efficient, accurate and very fast (~50 µs) parameterization of that cooling able to cope with CO₂ abundances from half the pre-industrial values to 10 times the current abundance.

Information on the mesospheric OH layer height is crucial for identifying sources of its variability and causes of discrepancies in measurements and models. Using space-based data, we inferred an empirical function for predicting the altitude of the layer at midlatitudes from ground-based measurements of OH intensity and temperature. By applying it to data at the Sierra Nevada Observatory, we found significant short-term variability in the layer altitude, mainly due to wave variability.

Noctilucent clouds (NLCs) are ice clouds that form at the polar summer mesopause and are very sensitive to temperature. They may therefore provide a way to monitor this remote region as our atmosphere changes. We show that temperature variations in the mesosphere are crucial for the growth of ice particles and that average fields are not enough to describe the process of NLC development. The paper also emphasises the difficulties in retrieving ice particle properties from optical observations.