1. Ability of global chemistry models to reproduce large scale UT/LS features of O3, CO, CH4 and acetone assessed by comparison to airborne CARIBIC data Frans Alkemade 1, Michiel van Weele 1, Peter van Velthoven 1, Sophie Szopa 2, Thierry Elias 2, Andreas Zahn 4, Tanja Schuck 3, Carl Brenninkmeijer 3 and the CARIBIC team (1) KNMI, Royal Netherlands Meteorological Institute, de Bilt, the Netherlands (2) Laboratoire des Sciences du Climat et de l’Environnement, CEA/CNRS/UVSQ,IPSL, Gif-sur-Yvette, France (3) Max Planck Institute for Chemistry, Atmospheric Chemistry Division, Mainz, Germany (4) Institute for Meteorology and Climate Research, Forschungszentrum Karlsruhe, Karlsruhe, Germany Presentation: The Extra-tropical UTLS: observations, concepts and future directions Community workshop at NCAR, Boulder, CO October 21, 2009

2. F. Alkemade (KNMI): presentation at NCAR UTLS workshop, October 21st 2009 Focus / Questions / Goals Can (our) models reproduce the observed large scale spatial and temporal variability and events? How can we use CARIBIC to evaluate several aspects of the atmospheric chemistry models (transport+chemistry)? Can we improve the models? (NB: No CARIBIC data analysis  see presentation Andreas Zahn)

3. F. Alkemade (KNMI): presentation at NCAR UTLS workshop, October 21st 2009 Overview - GEOMON: goal & means - CARIBIC: flights & data - TM5 : characteristics of model & emissions - LMDz-INCA : characteristics of model & emissions - Comparison CARIBIC observations to models (TM5 and LMDz INCA), years 2005-2006 (flights 113 – 174, Southern destinations only) O3 CH4 (from October 2005 onwards) CO Acetone (case study for March 2006) - Conclusions & Future Plans

4. F. Alkemade (KNMI): presentation at NCAR UTLS workshop, October 21st 2009 EU-GEOMON project (Global Earth Observation and Monitoring)  Pilot study on measurements usage. Partners: KNMI, Un. Oslo, CEA Some Objectives : Demonstrate the usage of Geomon Greenhouse gas observations (CH4, CO2) for global model evaluation Use CARIBIC observations to evaluate model simulated large scale gradients in the upper troposphere/lower stratosphere Compare the model simulations to a climatology based on airborne measurements of CO2 and CH4 from ~45 different aircraft campaigns.

5. F. Alkemade (KNMI): presentation at NCAR UTLS workshop, October 21st 2009 What is observed by Caribic? (Civil Aircraft for the Regular Investigation of the atmosphere Based on an Instrumented Container) In Flight: H2Ogas H2Ocloud Ozone NO NOy (the sum of all reactive nitrogen species: NO, NO2, HNO3, PAN, N2O5, HNO2, HNO4, and nitrates) CO In flight remote sensing: BrO, HCHO, OClO, O4 WAS (Whole Air Sample ; 12 canisters, analyzed in laboratory) : CO2, CH4, N2O, SF6 (+ others)

6. F. Alkemade (KNMI): presentation at NCAR UTLS workshop, October 21st 2009 CARIBIC flights

7. TM5 CTM some characteristics of model & emissions 2 x 3 grid, 34 layers ECMWF ERA Interim (EI) meteorology O3 : Nudging to top boundary condition (climatology based on GOME / SCIA observations) Emissions: selected datasets from GEMS and HYMN projects (see following slides) Methane is fixed in the lowest two layers to a single value (1760.0 ppb). Stratospheric loss by OH F. Alkemade (KNMI): presentation at NCAR UTLS workshop, October 21st 2009

8. Anthropogenic emissions for GEOMON Ship and aircraft emissions: EC 6 th FP IP Quantify Biomass burning emissions: GFEDv2 (van der Werf et al.) Other: year 2000 1x1 deg distributions IPCC/ACCENT; Remove ships and scale remainder with IIASA regional totals for 11 regions (‘MESSAGE’ land regions) For energy-related and non-energy related Year 2000 and scenarios for the years 2010, 2020,…2050 Linear interpolated emissions for the first decade: 2001,2002,…2009 Agricultural emissions from EDGAR (FT2000)

9. F. Alkemade (KNMI): presentation at NCAR UTLS workshop, October 21st 2009 Upscaling Asian CO emissions: AnthropogenicEDGAR, IIASA-scaled: 512 Tg(CO) NaturalPOET : 180 Tg(CO) Biomass burningGFEDv2 : 405 Tg(CO) Total for 2004: 1097 Tg(CO) increase of ~60% in China w.r.t IPCC/ACCENT (66 => 103 Tg) Asia CO emissions now about half the global total of non-biomass burning anthropogenic CO Scaling factors wrt IPCC/ACCENT

10. LMDz INCA (Laboratoire de Météorologie Dynamique - Interaction Chimie - Aérosols) characteristics of model & emissions LMDz-INCA is a GCM, u and v are nudged on ECMWF-interim analysis (other dynamical fields being computed by LMDz Emissions are those provided for the HYMN project (identical to those used by TM5) The model considers the chemistry of methane and NMHC (89 tracers). Stratospheric ozone is nudged to climate. For CH4 no fixed surface value is used (unlike TM5), therefore a bias is to be expected, due to an equilibrium problem (a long run with sources=sink should be necessary to stabilize the model at the correct concentrations). The reference papers describing INCA are Hauglustaine et al. JGR 2004 and Folberth et al. ACP 2005. F. Alkemade (KNMI): presentation at NCAR UTLS workshop, October 21st 2009

11. Example of model evaluation: the CARIBIC flights of February 2006 (138-141)

12. Reasonably good agreement for O3, CO, CH4 (Red = CARIBIC, Blue = TM5, Green = INCA) latitude altitude O3 CO CH4 Time 

13. Correlations between O3, CO, CH4 latitude altitude O3 CO CH4

14. O3 up, CO and CH4 down Correlations between O3, CO, CH4 O3 CO CH4

15. Correlations: O3 up, CO and CH4 down These features correlate with PV anomaly  stratospheric intrusion

16. Comparing Caribic – TM5 - LMDzINCA Profiles and scatter plots of O3 CO CH4 CO- O3 scatter plots: identifying chemical transition from UT to LS Acetone case study by LMDz INCA : 2006 flights #142-145 (Europe-South America). F. Alkemade (KNMI): presentation at NCAR UTLS workshop, October 21st 2009

17. Modelled vs Observed O3 Northern Hemisphere ( > 22) Dashed line: fit through all data Green line: data at locations where [O3] < 100 (troposphere) Blue line: data at locations where 100 < [O3] < 200 ( ~ UT/LS) Red line: data at locations where [O3] > 200 (stratosphere) Observations  Model  TM5 INCA

18. O3 vertical distribution NH ( > 22) TM5INCA Red: observations Blue: Model

19. TM5INCA Observations  Model  Scatter CH4 Northern Hemisphere ( > 22)

20. CH4 vertical distribution (all data) TM5INCA Red: observations Blue: Model

21. CH4 vertical distribution NH ( > 22) TM5 INCA Red: observations Blue: Model

22. CO Vertical distribution NH ( > 22) TM5 INCA Red: observations Blue: Model

23. Caribic INCA TM5 Intrusion branche is missing CO – O3 scatterplot : Modelled vs Observed Mixing Ratios

24. Acetone Comparison LMDz-INCA / CARIBIC #142 #145 #144 and #145 Branch of the acetone plume emitted by the Amazon forest North Atlantic Ocean Map of daily average of acetone concentration modeled by LMDz-INCA for 21 March 2006, with superimposed airborne CARIBIC measurements. Why acetone ? Acetone modifies radical concentrations and then intervenes in atmospheric chemistry: impacts on ozone cycle, greenhouse warming.

25. Future plans Test use of OD (operational) meteo data instead of EI (ERA Interim reanalysed) meteo. Improve upper boundary conditions (CH4, O3, CO) Look at more tracers (NO, NOy, etc.) More detailed assessment of GFED biomass burning and HYMN CH4 emissions. F. Alkemade (KNMI): presentation at NCAR UTLS workshop, October 21st 2009

26. Future: CH 4 anthropogenic and natural emissions HYMN CH 4 emissionsSource Global Totals in 2004 (Tg CH4) AnthropogenicEDGAR, IIASA-scaled293 Natural emissionsLPJ-based234 Biomass burningGFEDv2 vd Werff 20 Total emission547 Soil sinkLPJ-based-36 Net emission511 Anthropogenic Natural

27. Conclusions Variability / spread in the observations is not completely captured by models. (  problem with STE  better vertical model resolution needed in UT/LS) However, large scale features are captured by both models. Comparing the observed correlations of O3, CO and CH4 to the models can be used to assess the meteorological input fields. Direct comparison between CARIBIC observations and models may help to assess the quality of emissions/processes and of the boundary conditions used by the models. F. Alkemade (KNMI): presentation at NCAR UTLS workshop, October 21st 2009

28. Thank you

29. Some redundant slides follow F. Alkemade (KNMI): presentation at NCAR UTLS workshop, October 21st 2009

30. Anthropogenic NO X emissions: AnthropogenicEDGAR, IIASA-scaled26.5 TgN ShipQUANTIFY 4.4 TgN AircraftQUANTIFY 0.9 TgN NaturalPOET 8.0 TgN Biomass burningGFEDv2 5.3 TgN Lightningparameterised 5.4 TgN Total for 2004: 50.5 TgN Similar as used in IPCC 4AR/ACCENT study (IIASA regionally rescaled)

31. LMDz INCA characteristics of model & emissions LMDz-INCA is a GCM, u and v are nudged on ECMWF-interim analysis (other dynamical fields being computed by LMDz emissions are those provided for the HYMN project (the same than yours actually) It consider the chemistruy of methane and NMHC (89 tracers). We also nudged strato ozone on a climato. The reference papers describing INCA are Hauglustaine et al. JGR 2004 and Folberth et al. ACP 2005. F. Alkemade (KNMI): presentation at NCAR UTLS workshop, October 21st 2009

32. The CEA is the French Atomic Energy Commission (Commissariat à l'énergie atomique) *The National Institute of Sciences of the Universe **GEOMON is a scientific project within the 6th EU Framework Programme for Research and Development ***The instruments and networks will be integrated into a larger system, which includes space missions and the development of digital tools to ensure all the measurements are compatible. ****French laboratories involved in GEOMON: Laboratory of Climate and Environmental Sciences (IPSL/LSCE, a joint CEA, CNRS and Université Versailles Saint-Quentin laboratory), Aeronomics Service (IPSL/SA, a joint CNRS, Université Pierre et Marie Curie and Université Versailles Saint- Quentin laboratory), Laboratory of Dynamic Meteorology (IPSL/LMD, a joint CNRS, École polytechnique, Université Paris VI and ENS Paris laboratory), Inter-University Laboratory of Atmospheric Systems (LISA, a joint CNRS, Université Paris VII and Université Paris XII laboratory), the Laboratory of Atmospheric Optics (LOA, a joint Université Lille 1 and CNRS laboratory) and the Laboratory of the Atmosphere and Cyclones (LACy, a joint CNRS, Université La Réunion and Météo- France CNRM laboratory) F. Alkemade (KNMI): presentation at NCAR UTLS workshop, October 21st 2009

33. Tropical mid-troposphere O3 (-22 22 latitude) Observations  Model  TM5INCA

34. Comparing O3 (all data 2005 - 2006) Observations  Dashed line: fit through all data Green line: data at locations where [O3] < 100 (troposphere) Blue line: data at locations where 100 < [O3] < 200 ( ~ UT/LS) Red line: data at locations where [O3] > 200 (stratosphere) TM5INCA Model 

35. [CH4] < 1700 ppb  must be stratospheric intrusion (otherwise, air would originate before 1990; implausible) TM5