Sensitivity of Aeolus HLOS winds to temperature and pressure specification in the L2B processor

Abstract. The retrieval of wind from the first Doppler wind lidar of Europen Space Agency (ESA) launched in space in August 2018 is based on a series of corrections necessary to provide observations of a quality useful for Numerical Weather Prediction (NWP). In this paper we examine properties of the Rayleigh-Brillouin correction necessary for the retrieval of horizontal line-of-sight wind (HLOS) from a Fabry-Perot interferometer. This correction is taking into account the atmospheric stratification, namely temperature and pressure information that are provided by a NWP model as suggested prior launch. Since NWP models contain errors the main goal in the study is to evaluate the impact of these errors on the HLOS sensitivity by comparing the Integrated Forecast System (IFS) and Action de Recherche Petite Echelle Grande Echelle (ARPEGE) global model temperature and pressure short term forecasts collocated with the Aeolus orbit. The model error is currently not taken into account in the computation of the HLOS error estimate since its contribution is believed small. This study largely confirms this statement to be a valid assumption, although it also shows that model errors could locally (i.e. jet-stream regions, below 700 hPa over both earth poles and as well in stratosphere) be significant. For a future Aeolus follow-on missions this study suggests to consider realistic estimations of model errors in the HLOS retrieval algorithms, since this will lead to an improved estimation of the Rayleigh-Brillouin sensitivity uncertainty contributing to the HLOS error estimate and better exploitation of space lidar winds in NWP systems.