Mechanisms for varying non-LTE contributions to OH rotational temperatures from measurements and modelling. II. Kinetic model

OH rotational temperatures Trot deviate from kinetic temperatures Tkin if the rotational level populations derived from the measured OH lines are not in local thermodynamic equilibrium (LTE). In particular,. from OH bands wi Trot th high upper vibrational level appear to be affected by incomplete rotational relaxation and corresponding non-LTE temperature excesses ΔTNLTE. The processes that lead to these non-LTE effects and their variations have not been investigated in detail so far since this requires thorough modelling of the rotational relaxation based on suitable observing data. We performed such an effort for the vibrational level v=9. As it is the highest v that can be populated by the OH-producing hydrogen-ozone reaction, the relaxation processes are the least complex ones. Nevertheless, there are many systematic uncertainties. Therefore, we studied the ΔTNLTE depending on the Einstein A-coefficients used, the very uncertain rate coefficient for the v=9 deactivating collisions of OH with atomic oxygen, and the input profiles for Tkin, air density, and the volume mixing ratios of N2, O2, O, and O3. For the profiles, we used SABER products, where we adapted the O retrieval for a better agreement with the model, and the empirical NRLMSISE-00 model. The profiles were selected or calculated for the Cerro Paranal region in Chile since this allowed us to combine OH Trot measurements based on high-resolution spectra of the UVES spectrograph at the Very Large Telescope with the SABER Tkin and OH emission profiles for an alternative, fully observation-based ΔTNLTE derivation. Moreover, we measured OH(v=9) rotational level populations up to the rotational state N=10 based on UVES OH(9-3) and OH(9-4) line measurements. With these data in comparison with a grid of models, we derived the best-fitting crucial rate coefficients for the rotational relaxation at v=9, which have not been studied before. With the final models, we calculated ΔTNLTE climatologies and compared them with those from the UVES-based data. Significant ΔTNLTE and similar variability structures could be identified. Moreover, we analysed the model in order to find the mechanisms that cause the variations. Here, the steep increase of ΔTNLTE with altitude combined with the height variations of the OH layer appears to be the most important process.