Data composites of airborne observations of tropospheric ozone and its precursors

Tropospheric data from a number of aircraft campaigns have been gridded ont o global maps, forming "data composites" of chemical species important in o zone photochemistry. Although these are not climatologies in the sense of a long temporal average, these data summaries are useful for providing a pic ture of the global distributions of these species and are a start to creati ng observations-based climatologies. Using aircraft measurements from a num ber of campaigns, we have averaged observations of O-3, CO, NO, NOx, HNO3, PAN, H2O2, CH3OOH, HCHO, CH3COCH3, C2H6, and C3H8 onto a 5 degrees latitude by 5 degrees longitude horizontal grid with a 1-km vertical resolution. Th ese maps provide information about the distributions at various altitudes, but also clearly show that direct observations of the global troposphere ar e still very limited. A set of regions with 10 degrees-20 degrees horizonta l extent has also been chosen wherein there is sufficient data to study ver tical profiles, These profiles are particularly valuable for comparison wit h model results, especially when a full suite of chemical species can be co mpared simultaneously. The O-3 and NO climatologies generated from measurem ents obtained during commercial aircraft flights associated with the MOZAIC and NOXAR programs are incorporated with the data composites at 10-11 km. As an example of the utility of these data composites, observations are com pared to results from two global chemical transport models, MOZART and IMAG ES, to help identify incorrect emission sources, incorrect strength of conv ection, and missing chemistry in the models. These comparisons suggest that in MOZART the NO, biomass burning emissions may be too low and convection too weak and that the transport of ozone from the stratosphere in IMAGES is too great. The ozone profiles from the data composites are compared with o zonesonde climatologies and show that in some cases the aircraft data agree with the long-term averages, but in others, such as in the western Pacific during PEM-Tropics-A, agreement is lacking. Finally, the data composites p rovide temporal and spatial information, which can help identify the locati ons and seasons where new measurements would be most valuable. All of the d ata composites presented here are available via the Internet (http://aoss.e ngin.umich.edu/SASSarchive/).