1. Vertical transport of chemical compounds from the surface to the UT/LS: What do we learn from SEAC 4 RS? Qing Liang 1 & Thomas Hanisco 2, Steve Wofsy 3, Jasna Pittman 3, Elliot Atlas 4, Sue Schauffler 5, Donald R. Blake 6, Rushan Gao 7, Thomas Ryerson 7, Paul Wennberg 8, Michael Manyin 9 1 NASA Goddard Space Flight Center/ USRA GESTAR 2 NASA Goddard Space Flight Center 3 Harvard University 4 University of Miami 5 NCAR Atmospheric Chemistry Observations and Modeling Laboratory 6 University of California, Irvine 7 NOAA ESRL 8 California Institute of Technology 9 NASA Goddard Space Flight Center/SSAI

2. Motivation and Goals Surface to the UT/LS transport of trace gases can exert significant impacts on atmospheric abundances of ozone (O 3 ) and hydroxyl radical (OH), which will have chemical and radiative impacts on the atmosphere. Vertical transport of CO and NMHCs Vertical transport of very-short-lived (VSL) halogen compounds, i.e. CHBr 3 and CH 2 Br 2. Questions of interest What is the surface to UT/LS transport timescale? What are transport efficiency of chemical compounds (the impact of tracer lifetime and solubility)? What are the regional differences in convective lofting – North America (SEAC 4 RS & DC 3 ) vs. the Western Pacific (ATTREX/CONTRAST)?

3. V2 - GEOS-5 with standard stratospheric chemistry and detailed VSL bromocarbons V3 - GEOS-5 with stratosphere-troposphere full chemistry GCM simulations for the SEAC 4 RS period o Horizontal resolution is 2° latitude x 2.5° longitude o 72 vertical layers from surface to 0.01 hPa, with ~1 km/layer resolution at the UT/LS region (will conduct a future simulation with MERRA meteorology further comparison) NASA GEOS-5 Chemistry Climate Model (GEOSCCM) The model simulations are sampled at the same geographical and vertical location for the corresponding time as the observations for model-observation comparison. Two new CO 2 tracers – transport clock as well as model transport diagnostics 1 ) Regular CO 2 - with surface boundary conditions from the NOAA observations (latitude and monthly variable) 2) Annual CO 2 – same as regular CO 2 but with seasonal cycle removed at all latitude grids.

4. Model annual-based CO 2 shows a linear correlation with age of air, thus an accurate clock tracer for transport CO 2 as a transport clock tracer

5. However, the actual application as a clock tracer is complicated due to the surface seasonal variations in CO 2 Model CO 2 with seasonality

6. Evaluation of model transport with observed CO 2 GEOSCCM modeled CO 2 agrees well with the SEAC 4 RS observations A slight high CO 2 bias at 350-370 K – vertical transport is biased slow (insufficient convective injection?)

7. Quantification of vertical transport timescale using model age of air Average ascent rate ~ 6 days/K (2 months/10K) Model transport slow-biased?

8. Surface to UT/LS transport of pollution tracers COC2H2C2H2 C2H6C2H6 Loss reactionsCO + OH ->C 2 H 2 + OH -> C 2 H 2 + O 3 -> C 2 H 6 + OH -> C 2 H 5 --> CH 3 CHO --> PAN C 2 H 6 + Cl -> C 2 H 5 --> CH 3 CHO --> PAN 400-450K~ 2 years~ 6 months~ 4 months 350-400K~ 1 year~ 10 months Troposphere~ 1 month~ 0.5 month~ 3 months Tracer lifetime At different layers The increasing efficiency of C 2 H 6 loss via the C 2 H 6 + Cl reaction creates a large vertical as well as tropical vs. high-latitude gradients in C 2 H 6 -> making C 2 H 6 an excellent transport tracer for tropical convective lofting. Convective injection above 350K?

9. GEOSCCM vs. Observations Credible transport Credible OH Lack of the critical C 2 H 6 + Cl; Missing surface emissions Why a stratospheric ozone bias (~ +200 ppb)? Chemical? Not enough convective scavenging

10. Surface to UT/LS transport of CHBr 3 and CH 2 Br 2 -- Model vs. Observations CHBr 3 CH 2 Br 2 Model high bias (JPL-10 k-OH-CH 2 Br 2 too slow for T ~200K range??) Tropospheric low biases -> Surface emissions of CHBr 3 and CH 2 Br 2 are too small? This contradicts the findings from the Western Pacific observations (CONTRAST) as well as those in Hossaini et al. (2013) (GEOSCCM emissions are fine, slightly high-based).

11. High levels of CHBr 3 and CH 2 Br 2 during SEAC 4 RS SEAC 4 RS ~12 ppt Br in the MBL

12. High CHBr 3 /CH 2 Br 2 ratios during SEAC 4 RS Pacific campaigns Emission ratio: 1.3 ppt CHBr 3 /1 ppt CH 2 Br 2

13. High CHBr 3 /CH 2 Br 2 ratios during SEAC 4 RS Pacific campaigns Emission ratios: 1.3 ppt CHBr 3 /1 ppt CH 2 Br 2 3.0 ppt CHBr 3 /1 ppt CH 2 Br 2 (Gulf of Mexico)

14. High CHBr 3 /CH 2 Br 2 ratios during SEAC 4 RS Pacific campaigns Emission ratios: 1.3 ppt CHBr 3 /1 ppt CH 2 Br 2 3.0 ppt CHBr 3 /1 ppt CH 2 Br 2 2.0 ppt CHBr 3 /1 ppt CH 2 Br 2 (Gulf of Mexico) (US East coast) The Atlantic Ocean features much higher biogenic bromocarbon emissions, in particular CHBr 3, than the Pacific Ocean

15. ~ 6 ppt organic Br at the base of TTL + inorganic products (ppt??) VSL bromine at the North American tropical UT

16. Rapid convective ventilation below 355K topped by slow ascent from 355K to 450K at a rate of ~6 days/K Among all potential choices, C 2 H 6 appears as an excellent convective transport tracer for pollution SEAC4RS C 2 H 6 observations indicate sporadic convective injection above 355K The Golf of Mexico (and the Atlantic Ocean) features very high emissions of very-short-lived bromocarbons, in particular CHBr 3 (~12 ppt organic Br available at the surface) – Need to revisit the VSLS emissions as the same emission distribution does apply to both CHBr 3 and CH 2 Br 2 – How much of this surface VSL bromine gets to the stratosphere? How does the surface to UT/LS transport over continental US (DC 3 ) and the western Pacific (CONTRAST/ATTREX) differ from SEAC 4 RS? Conclusions and future work