1. Modeling deep convection in the oceanic tropical region C. Mari Laboratoire d'Aérologie, CNRS, France SPARC-GEWEX-IGAC joint meeting Victoria Island, June 2006

2. From B. Stevens, Annu. Rev. Earth. Planet. Sci, 2005 Cloud regimes in thermally direct circulations. Adapted from Arakawa (1975) TTL

3. Oceans Differences between the Atlantic and Pacific ITCZ: position, intensity, and seasonal cycle in the ITCZ; SST gradient; land effects; remote influences

4. Trimodal distribution of convection and cumulus congestus Many conceptual models of tropical convection are based on a BIMODAL cloud distribution, emphasizing shallow “trade- wind” or boundary layer cumuli and deep cumulonimbi. TOGA COARE results emphasize the dominance of cumulus congestus and point to a TRIMODAL cloud distribution in which the freezing level inversion is the key From Johnson et al. 1999, J. Clim.

5. Role of entrainment and detrainment: Deep convective clouds (over tropical oceans) are not undilute hot towers... ε δ From Lawrence and Rasch, JAS, 2005 Bulk formulations may underestimate transport in the upper troposphere for tracers with lifetime of a couple of days or less !

6. Diurnal cycle of oceanic convection From Tian et al., JGR, 2004 -> Nocturnal-early morning convection maximum (0600 LT) occurs over open oceans -> UTH lags high cloud peak (around 1600 LT) by 6-8 hours -> Physical mechanisms behind the diurnal cycle of oceanic convective clouds poorly understood (diurnal cycle of SST ?) -> Model fails to reproduce the magnitude and diurnal phase of the UTH -> nocturnal vs. diurnal chemistry ?

7. Ozone at 20°S MIR SAOZ March 2001 % deviation from zonal mean Minima in TTL over convective areas, particularly maritime (O3 destruction at surface level over Ocean), Maxima over subsident regions (meridional horizontal transport) Ozone distribution mainly controlled by transport (quasi-horizontal from mid-latitude and vertical by convection) -10 -20 -50 -10 30 40 20 5 5 Z 360-340 O3 deviation (%) From Pommereau et al., ESA Role of the convective transport Ind. Ocean Pacific OceanAtl. Ocean

8. Role of the convective transport From Solomon et al., GRL, 2005 Observed reduced or near-zero ozone in the tropics: --> Primary role of deep convection (max. over southwest pacific region) --> Secondary role of horizontal transport (?) and chemistry (cirrus/aerosols) deep layer

9. Term 1 Term2 Term3Term4 k 1 [NO][HO 2 ] – {k 4 [O( 1 D)][H 2 O] + k 6 [HO 2 ][O 3 ]+ k 5 [OH][O 3 ] } Production (O 3 ) – Loss (O 3 ) Term1 accounts for ~ 80% of O 3 Production Net O 3 depends upon HO x (OH + HO 2 ) Net O 3 depends upon NO x (NO + NO 2 ) Net O 3 depends upon O 3 photolysis which produces O( 1 D) Net O 3 depends upon H 2 O (i.e., H 2 O is a reactant and not just a climatic gas) Net O 3 depends upon peroxides, and OVOC which produce HO x when photolyzed In UT get O 3 production with increasing HO x (Term 1) In LS get O 3 destruction with increasing HO x (Terms 3 & 4) In UT/LS get a mixture -> good testing of models ! Role of the chemical processes How does convection perturb the chemistry regimes in the UTLS? Are the chemistry regimes/rates significantly changed at low temperatures?

10. Additional Sources of HO x in the UT/LS Low P(HOx) High NOx High P(HOx) Low NOx Low P(HOx) Low NOx High P(HOx) High NOx role of CH3OOH as HOx precursor --> Role of water vapour and peroxydes as HOx precursors --> No unique signature of oceanic convection / large-scale influence From Mari et al. (2003) NASA/PEM-TROPICS B

11. Short lived halocarbons(<0.5y) CHBr 3, CH 2 Br 2,CH 3 I,... From Salawitch et al. (Nature, 2006) Role in catalytic destruction of ozone in the stratosphere AND troposphere Current effort to refine the global distribution of oceanic surface fluxes Large emissions over open ocean where deep convection takes place.

12. Loss by photolysis (~4 days), large aircraft data base [D.R. Blake, UCI] Cl - Emission h Simple model for ocean source DOC CH 3 I(aq) Methyl iodide (CH 3 I): potential tracer of marine convection

13. Need to improve definition of source! Model vs. observed CH 3 I(aq) field (r 2 =0.40)

14. Lower electrical activity over ocean LIS LIGHTNING FLASHES (2000) DJF JJA

15. Martin et al. [2002] Difference between standard and no-lightning GEOS-CHEM simulations... but influence of continental lightning on ozone over the ocean

16. Indian Ocean Alt Ocean SPCZ West Pacific MLS data, Spichtinger et al., QJRMS, 2003 Role of cirrus on NOx budget in the UTLS Important factor in the dehydration of air --> changes in H 2 O --> changes in HOx --> changes in O 3. – heterogeneous chemistry ice surfaces. Contribution to ozone destruction in the UTLS. How could we represent this sink in CRM and global models properly(not at equilibrium)? Karcher and Voigt, GRL, 2006 decreasing T increasing fraction of HNO3 in ice

17. Are deep convective clouds efficient aerosol production engines (over the ocean)? T -50°C -25°C 0°C Soluble Partly Soluble (SO 2 ) Insoluble Homogeneous nucleation H 2 SO 4 Condensational growth From Kulmala et al., submitted to JGR, 2006 Hermann et al., JGR, 2003: identify ITCZ as a region of pronounced particle formation Lee et al., JGR, 2004: particles formation observed in cirrus clouds,... Atmospheric nucleation pathways have not been resolved. Several mechanisms are suggested: ion mediated, binary, ternary or nucleation of water insoluble vapors (organic compounds) Saturation ratios in UT >> in the BL !!

18. Challenges... Convective upward and downward transports Entrainment of environmental air and multilevel detrainment Better representation of the structured systems + associated stratiform precipitations Diurnal cycle of oceanic convection vs. diurnal cycle of chemical species Role of VSL Halocarbons (Oceanic emissions) “Tracers” of marine convection (CH3I, ozone, ???) HOx precursors over oceanic regions: acetone, CH3OOH, CH2O, H2O2,...??? Lightning and large-scale influence from the continents Role of cirrus clouds (sink for HNO3 – satellite measurements MLS, IASI...) Budget and chemical composition of UT aerosols -> need for explicit simulations for chemical-microphysical-dynamical coupling -> need for long-term high-resolution simulations (impact of local chemistry on ozone,...) -> use of satellite measurements over remote oceanic regions

19. From Liu and Zipser, JGR, 2005 Does convection penetrate the tropical tropopause over the oceans? 20°N 20°S 20°N 20°S 20°N 20°S 20°N 20°S 20°N 20°S Overshooting Precipitation Features using different reference heights. West Pacific SPCZ Atl ITCZ Indian ocean

20. From Wennberg et al., Science, Vol. 279, 49-53, 1998. Additional Sources of HO x in the UT/LS --> Role of acetone as HOx precursor

21. a priori sources/sinks Optimized sources/sinks (including “microbial” ocean sink, photochemical ocean source) Additional Sources of HO x in the UT/LS: Oceanic signature in atmospheric acetone observations southern Sweden [Solberg et al., 1996] South Pacific [Singh et al., 2001] --> measurements of oceanic acetone are very few! From Singh et al. (SOLAS)