Hydroxyl layer: Mean state and trends at midlatitudes

Based on an advanced model of excited hydroxyl relaxation we calculate trends of number densities and altitudes of the OH*‐layer during the period 1961–2009. The OH*‐model takes into account all major chemical processes such as the production by H + O 3 , deactivation by O, O 2 , and N 2 , spontaneous emission, and removal by chemical reactions. The OH*‐model is coupled with a chemistry‐transport model (CTM). The dynamical part (Leibniz Institute Model of the Atmosphere, LIMA) adapts ECMWF/ERA‐40 data in the troposphere‐stratosphere. The change of greenhouse gases (GHGs) such as CH 4 , CO 2 , O 3 , and N 2 O is parameterized in LIMA/CTM. The downward shift of the OH*‐layer in geometrical altitudes occurs entirely due to shrinking (mainly in the mesosphere) as a result of cooling by increasing CO 2 concentrations. In order to identify the direct chemical effect of GHG changes on OH*‐trends under variable solar cycle conditions, we consider three cases: (a) variable GHG and Lyman‐α fluxes, (b) variable GHG and constant Lyman‐α fluxes, and (c) constant GHG and Lyman‐α. At midlatitudes, shrinking of the middle atmosphere descends the OH*‐layer by ~ −300 m/decade in all seasons. The direct chemical impact of GHG emission lifts up the OH*‐layer by ~15–25 m/decade depending on season. Trends of the thermal and dynamical state within the layer lead to a trend of OH* height by ~ ±100 m/decade, depending on latitude and season. Trends in layer altitudes lead to differences between temperature trends within the layer, at constant pressure, and at constant altitude, respectively, of typically 0.5 to 1 K/decade.