1Institut d’Aéronomie Spatiale de Belgique (IASB-BIRA), Brussels, Belgium
2Institute for Environmental Physics, University of Heidelberg, Germany
3Service d’Aéronomie du CNRS, Verrieres le Buisson, France
*now at: Service de Chimie Quantique et Photophysique: atomes, molécules et atmosphères, Université Libre de Bruxelles, Belgium
**now at: Jet Propulsion Laboratory, Pasadena, California, USA
Abstract. A retrieval algorithm based on the Optimal Estimation Method (OEM) has been developed in order to provide vertical distributions of NO2 in the stratosphere from ground-based (GB) zenith-sky UV-visible observations. It has been applied to observational data sets from the NDSC (Network for Detection of Stratospheric Change) stations of Harestua (60° N, 10° E) and Andøya (69° N, 16° E) in Norway. The information content and retrieval errors have been analyzed following a formalism used for characterizing ozone profiles retrieved from solar infrared absorption spectra. In order to validate the technique, the retrieved NO2 vertical profiles and columns have been compared to correlative balloon and satellite observations. Such extensive validation of the profile and column retrievals was not reported in previously published work on the profiling from GB UV-visible measurements. A good agreement - generally better than 25% - has been found with the SAOZ (Système d'Analyse par Observations Zénithales) and DOAS (Differential Optical Absorption Spectroscopy) balloons. A similar agreement has been reached with correlative satellite data from the HALogen Occultation Experiment (HALOE) and Polar Ozone and Aerosol Measurement (POAM) III instruments above 25km of altitude. Below 25km, a systematic underestimation - by up to 40% in some cases - of both HALOE and POAM III profiles by our GB profile retrievals has been observed, pointing out more likely a limitation of both satellite instruments at these altitudes. We have concluded that our study strengthens our confidence in the reliability of the retrieval of vertical distribution information from GB UV-visible observations and offers new perspectives in the use of GB UV-visible network data for validation purposes.