1Harvard University, Department of Earth and Planetary Sciences, Cambridge, MA, USA
2Harvard University, School of Engineering and Applied Sciences, Cambridge, MA, USA
3Harvard-Smithsonian Center for Astrophysics, Cambridge, MA, USA
4University of Maryland, College Park, College of Computer, Mathematical and Natural Science, College Park, MD, USA
5University of Maryland, College Park, Department of Atmospheric and Oceanic Science, MD, USA
6NASA Goddard Space Flight Center, Greenbelt, MD, USA
7Laboratoire d'Aerologie, UMR5560, CNRS and Universite de Toulouse, Toulouse, France
Abstract. We present a global data set of free tropospheric ozone–CO correlations with 2° × 2.5° spatial resolution from the Ozone Monitoring Instrument (OMI) and Atmospheric Infrared Sounder (AIRS) satellite instruments for each season of 2008. OMI and AIRS have near-daily global coverage of ozone and CO respectively and observe coincident scenes with similar vertical sensitivities. The resulting ozone–CO correlations are highly statistically significant (positive or negative) in most regions of the world, and are less noisy than previous satellite-based studies that used sparser data. Comparison with ozone–CO correlations and regression slopes (dO3/dCO) from MOZAIC (Measurements of OZone, water vapour, carbon monoxide and nitrogen oxides by in-service AIrbus airCraft) aircraft profiles shows good general agreement. We interpret the observed ozone–CO correlations with the GEOS (Goddard Earth Observing System)-Chem chemical transport model to infer constraints on ozone sources. Driving GEOS-Chem with different meteorological fields generally shows consistent ozone–CO correlation patterns, except in some tropical regions where the correlations are strongly sensitive to model transport error associated with deep convection. GEOS-Chem reproduces the general structure of the observed ozone–CO correlations and regression slopes, although there are some large regional discrepancies. We examine the model sensitivity of dO3/dCO to different ozone sources (combustion, biosphere, stratosphere, and lightning NOx) by correlating the ozone change from that source to CO from the standard simulation. The model reproduces the observed positive dO3/dCO in the extratropical Northern Hemisphere in spring–summer, driven by combustion sources. Stratospheric influence there is also associated with a positive dO3/dCO because of the interweaving of stratospheric downwelling with continental outflow. The well-known ozone maximum over the tropical South Atlantic is associated with negative dO3/dCO in the observations; this feature is reproduced in GEOS-Chem and supports a dominant contribution from lightning to the ozone maximum. A major model discrepancy is found over the northeastern Pacific in summer–fall where dO3/dCO is positive in the observations but negative in the model, for all ozone sources. We suggest that this reflects a model overestimate of lightning at northern midlatitudes combined with an underestimate of the East Asian CO source.