kim68@fas.harvard.edu

Global Ozone-CO Correlations from OMI & AIRS as Constraints on Ozone Sources and Transport

Patrick Kim1, Daniel Jacob1, Xiong Liu2, and Juying Warner3 1Harvard University, Department of Earth and Planetary Science | 2Harvard-Smithsonian Center for Astrophysics | 3University of Maryland, College Park

Summary & Conclusions

Sensitivity to Transport Error Motivation 2008 Ozone-CO Correlations Current global models can capture the observed large-scale spatial and seasonal patterns of ozone concentrations but there is large uncertainty in the driving factors, as reflected by the large differences between models in ozone production and loss rates (Wu et al., 2007) and in source contributions (Fiore et al., 2009). Ozone-CO correlations have long been used as in-situ constraints on ozone sources and transport, but have only recently become available from satellite retrievals (Zhang et al., 2006). All previous ozone-CO correlation analyses from space have used ozone satellite data from the TES instrument. A drawback is the limited coverage of TES – individual satellite retrievals can have large random error, compromising correlation analyses unless large ensembles are considered. We present here a new high-density global satellite database of ozone-CO correlations using data from the Ozone Monitoring Instrument (OMI) and the Atmospheric Infrared Sounder (AIRS) instruments and explore its value for constraining our understanding of the factors controlling ozone.

We average the individual OMI ozone and AIRS CO profile retrievals over a 2° x 2.5° horizontal grid for each day of 2008. We focus on the 700-400 hPa columns, where both instruments have maximum resolution. These are computed by adding the corresponding partial columns in the retrievals, with linear interpolation as necessary. We compute seasonal ozone-CO correlation statistics for 3-month time series (DJF, MAM, JJA, SON) of the ozone and CO mixing ratios for each grid square. For comparison to the satellite data, we archive GEOS-Chem daily output at the local satellite overpass time and regrid vertically to the instrument retrieval levels. The simulated ozone and CO profiles are smoothed with the instrument averaging kernels and the statistics are calculated using the exact same methodology as the satellite observations.

•  We developed a new global database of ozone-CO correlations and showed its value as a constraint on ozone sources with the GEOS-Chem CTM

•  Tests applying random retrieval error to the GEOS-Chem profiles indicate negligible effect of retrieval noise on the simulated ozone-CO correlations.

•  The impact of ozone sources on the ozone-CO relationship is quantified using Δ and we calculate regional statistics for areas of interest. For example, we confirm large contributions from lightning and stratospheric intrusions on the South Atlantic maximum during DJF.

•  We use the difference between the simulated ozone-CO relationship when GEOS-Chem is driven by GEOS-4 and GEOS-5 as a proxy for transport error and use it to diagnose areas where the ozone-CO correlations are not useful for quantifying ozone sources.

•  We are currently exploring the utility of the observed ozone-CO correlations as a constraint on excessive lightning emissions in the model and will explore interannual variability in future work.

References: Chin et al. (1994), J. Geophys. Res., 99, 14565-14573. Fiore et al. (2009), J. Geophys. Res., 114, D04301. Voulgarakis et al. (2011), Atmos. Chem. Phys., 11, 5819-5838. Wu et al. (2007), J. Geophys. Res., 112, D05302. Zhang et al. (2006), Geophys. Res. Lett., 33, L18804. This research is supported in part by the Department of Energy Office of Science Graduate Fellowship Program (DOE SCGF), made possible in part by the American Recovery and Reinvestment Act of 2009, administered by ORISE-ORAU under contract no. DE-AC05-06OR23100.

Ozone-CO correlations reflect a combination of transport and ozone chemistry. In general, however, the relative contributions of transport and chemistry in driving the ozone-CO relationship is not obvious, especially in the free troposphere where the chemistry is relatively slow and pollution enhancements are relatively weak. These complications motivate the use of a chemical transport model (CTM) to interpret the observed ozone-CO relationship. We use the global GEOS-Chem CTM (version 9-01-02) for this purpose with a focus on ozone sources below. Zhang et al. (2006) and Voulgarakis et al. (2010) previously diagnosed the influence of individual sources on the ozone-CO relationship as the change in the ozone-CO slope between their base CTM simulation and a sensitivity simulation with that source shut off. However, this does not provide a proper measure of the influence of the source on ozone because the ozone-CO slope is then affected by changes in both ozone and CO. A more appropriate approach is to correlate CO from the base simulation with the ozone difference between the base and sensitivity simulations. Chin et al. (1994) used this approach to interpret ozone-CO relationships from North American surface sites in terms of ozone production. We illustrate this in an example grid box receiving polluted outflow from North America below and apply the approach in the following sections.

GEOS-Chem is driven by assimilated meteorological fields provided by NASA GMAO. Here we examine the impact of model errors in transport on the simulated ozone-CO relationship by comparing results from a simulation driven by the latest version of the GEOS fields, GEOS-5, and its predecessor GEOS-4. The two fields are very different in terms of both model physics and data assimilated. A key difference is the parameterization for convection used in the two versions - due to the sub-grid nature of convection, this process must be parameterized. Regions highly sensitive to the choice of assimilated meteorology highlight where transport in the model limits the ability to interpret ozone sources.

Interpreting the O3-CO Relationship

Regional Analysis

OMI/AIRS 2006 Combustion Sources

Biogenic Sources

Stratosphere

Lightning NOx

OMI/AIRS 2008

w/ GEOS-4 2006

w/ GEOS-5 2006

w/ GEOS-5 2008

GEOS-Chem (GC) GC without

LEGEND

JJA

West Pacific (20-40 °N, 120-160 °E)

North Atlantic (30-40 °N, 60-80 °W)

JJA

East Pacific (0-20 °S, 80-120 °W)

SON

South Atlantic (0-20 °S, 30 °W-10 °E)

DJF

We calculate ozone-CO RMA regression slope statistics for regions of interest below. Error bars show 95% bootstrapped confidence intervals.

CO (ppbv)

O3 (

ppbv

)

ΔO

3 (pp

bv)

Simulation without Combustion Sources

Baseline Simulation

Ozone-CO Relationship at 38° N, 70° W

(Left) Ozone-CO relationship at 700-400 hPa in June-July-August (JJA) for a 2o x 2.5o grid square over the western North Atlantic. Values are GEOS-Chem model results sampled daily at the 13:30 local time of the OMI/AIRS satellite overpass. Ozone concentrations are from the baseline simulation (black) and for the sensitivity simulation without combustion sources (red). CO concentrations are from the baseline simulation. (Right) Difference in ozone between the baseline and sensitivity simulations as a function of CO concentrations from the baseline simulation. Δ is the reduced major axis (RMA) regression slope of this relationship, which measures the effect of the combustion source of ozone on the ozone-CO relationship. The range in parentheses is the nonparametric 95% bootstrapped confidence interval.

Ozone-CO correlation coefficients for OMI/AIRS and GEOS-Chem for each season of 2008. The correlation coefficients are computed from daily data on the 2o x 2.5o GEOS-Chem grid. Gray indicates missing data.

Based on the comparisons above, we examine here whether the ozone-CO relationship as observed from space can be used to place constraints on combustion, biogenic, lightning, and stratospheric sources of ozone. All four are recognized to be major sources of ozone, but their relative importance in different regions of the troposphere is uncertain. To this end, we apply the methodology for interpreting the slope (section to the left) for a number of regions during different seasons (section to the right). •  GEOS-Chem shows broad skill in reproducing the observed ozone-CO relationship in

continental outflow (North Atlantic & West Pacific) and in the subsiding branch of the Walker circulation (South Atlantic). This gives us confidence in the ozone source breakdown shown by the Δ of the sensitivity simulations.

•  While GEOS-Chem underestimates the magnitude of the ozone-CO relationship over the East Pacific when driven by GEOS-4 meteorology, the model consistently gets the wrong sign when driven by GEOS-5 meteorology. We attribute this difference in model ability to differences in transport rather than differences in natural emissions.

= d[ΔO3]/d[CO] = 0.32 (0.26 – 0.41) Δ

Ozone-CO correlations for JJA 2006 from OMI/AIRS, GEOS-Chem driven by GEOS-4 met fields, and GEOS-Chem driven by GEOS-5 met fields.

Α530-0435