1Chalmers University of Technology, Göteborg, Sweden
2Institute of Astrophysics and Geophysics, University of Liège, Liège, Belgium
3Institute of Environmental Physics, University of Bremen, Bremen, Germany
4Karlsruhe Institute of Technology (KIT), Institute for Meteorology and Climate Research (IMK-ASF), Garmisch-Partenkirchen, Germany
5Karlsruhe Institute of Technology (KIT), Institute for Meteorology and Climate Research (IMK-ASF), Karlsruhe, Germany
Abstract. A multiple regression model has been used to estimate linear trends of the CH4 and N2O total columns measured with the ground-based solar FTIR technique at four European stations, i.e. Jungfraujoch (47° N, 8° E, 3600 m a.s.l.), Zugspitze (47° N, 11° E, 3000 m a.s.l.), Harestua (60° N, 11° E, 600 m a.s.l.) and Kiruna (68° N, 20° E, 400 m a.s.l.). The total columns were retrieved with a common method developed within the EU-project HYMN. Anomalies from air pressure, total columns of hydrogen fluoride (HF) and carbon monoxide (CO) and tropopause height were used in the regression model to reduce the time series variability and thereby estimate trustful trends. Significant positive CH4 trends of 0.13–0.25% yr−1 at the 2-σ level were found for all participating stations for the 1996–2009 period. The strongest trends were estimated at northern latitudes stations while slightly weaker trends were observed in the Alps. For the time period of 2007–2009 a strong increase in the CH4 total column was observed for all stations with the strongest yearly growth at Kiruna (1.15 ± 0.17% yr−1). Significant positive N2O trends of 0.19–0.40% yr−1 were found for all stations in the 1996–2007 period with the strongest trend at Harestua. From the N2O data also crude tropospheric and stratospheric partial columns were derived, indicating that the observed difference in the N2O trends between the FTIR sites is of stratospheric origin. This agrees well with the N2O measurements by the Odin/SMR satellite showing the highest trends at Harestua 0.98 ± 0.28% yr−1, and considerably smaller trends in the alp regions 0.27 ± 0.25% yr−1. The multiple regression model was compared with two other trend methods, the ordinary linear regression and a Bootstrap algorithm. The multiple regression model estimated CH4 and N2O trends that differed by 12–31% compared to the other two methods. Since the trends estimated with the multiple regression model were carefully validated this stresses the importance to account for the atmospheric variability when estimating trends of CH4 and N2O total columns.