1Institute of Applied Physics, University of Bern, Switzerland
2MeteoSwiss, Payerne, Switzerland
3Jet Propulsion Laboratory, California Institute of Technology, Pasadena, USA
4University of Massachusetts, USA
5Institut für Meteorologie und Klimaforschung, Forschungszentrum Karlsruhe und Universität Karlsruhe, Germany
6Department of Atmospheric Physics, St. Petersburg State University, Russia
7Finnish Meteorological Institute, Helsinki, Finland
Abstract. Stratospheric O3 profiles obtained by the satellite limb sounders Aura/MLS, ENVISAT/MIPAS, ENVISAT/GOMOS, SAGE-II, SAGE-III, UARS/HALOE are compared to coincident O3 profiles of the ground-based microwave radiometer SOMORA in Switzerland. Data from the various measurement techniques are within 10% at altitudes below 45 km. At altitudes 45–60 km, the relative O3 differences are within a range of 50% Larger deviations at upper altitudes are attributed to larger relative measurement errors caused by lower O3 concentrations. The spatiotemporal characteristics of the O3 differences (satellite – ground station) are investigated by analyzing about 5000 coincident profile pairs of Aura/MLS (retrieval version 1.5) and SOMORA. The probability density function of the O3 differences is represented by a Gaussian normal distribution (except for profile pairs around the stratopause at noon). The dependence of the O3 differences on the horizontal distance between the sounding volumes of Aura/MLS and SOMORA is derived. While the mean bias (Aura/MLS – SOMORA) is constant with increasing horizontal distance (up to 800 km), the standard deviation of the O3 differences increases from around 8 to 12% in the mid-stratosphere. Geographical maps yield azimuthal dependences and horizontal gradients of the O3 difference field around the SOMORA ground station. Coherent oscillations of O3 are present in the time series of Aura/MLS and SOMORA (e.g., due to traveling planetary waves). Ground- and space-based measurements often complement one another. We introduce the double differencing technique which allows both the cross-validation of two satellites by means of a ground station and the cross-validation of distant ground stations by means of one satellite. Temporal atmospheric noise in the geographical ozone map over Payerne is significantly reduced by combination of the data from SOMORA and Aura/MLS. These analyses illustrate the synergy between ground-based and space-based measurements.