Tropospheric Chemistry Overview (or, 40 years in 20 minutes) Jennifer A. Logan Recent Results in Planetary Sciences, Atmospheric Chemistry, Climate and Energy Policy A symposium in celebration of Michael McElroy's contributions March 20, 2010.

A short history of tropospheric chemistry highlighting MBM’s contributions  1949: Migeotte - identifies CO in the atmosphere in solar absorption, ~100 ppb  1969: Weinstock – shows lifetime of CO is ~1 month, based on 14 CO and its budget, and suggests removal by OH  1971: Levy - first model of tropospheric OH, ~3x10 6 molec/cm 3 at noon; CO lifetime is ~2 months, and CH 4 oxidation is a source of HCHO (15 reactions)  1971, 1972: McConnell, McElroy, and Wofsy - show that CH 4 oxidation is a large source of CO (also suggest sources from terpenes) (1-d model, 32 rxns) Also show that CH 4 oxidation is a source of H 2 O in the stratosphere  1973: Chameides and Walker, Crutzen - first models of tropospheric ozone

Science, Levy Nature, McConnell et al. CH 4 data used to test model

 Mid-late 1970’s – the stratosphere and Mars  1975: Two postdocs join MBM’s group

 1981: Logan, Prather, Wofsy and McElroy - global model of trop. chemistry, constrained by observations of O 3, CO, CH 4, HNO 3, H 2 O (51 rxns) inventory for all sources of CO (P=2740 Tg) remote sources of NO x are ~10 Tg, based on almost no data used CH 3 CCl 3 (MCF) and a box model to test OH used OH to infer global budget for CH 4 (580 Tg) and other gases trop. ozone budget, P=3840 Tg, L=2820 Tg  All within ~20% of present day values

CO observations OH, HO 2 vs. NO NO x observations and model (lines)

Atmospheric Chemistry within a General Circulation Model  : Mahlman & Moxim, Levy et al. - first tracer model of ozone using archived GCM fields, no chemistry  1981: McElroy and Wofsy propose to NASA to develop “a global 3-D model with realistic dynamics and chemistry”, using the GCM developed by Jim Hansen et al. at GISS, building on their tracer model (Section 4 of a 103 page proposal!) Issues identified include an accurate method for transport tracer conservation efficient techniques for chemical rates use of activity data for anthropogenic emissions

Strategy outlined in the 1981 proposal Elucidate transport mechanisms, and include a strategy for model validation using observations 1.Use CFCs to test interhemispheric transport  Simple chemistry  Requires strict numerical accuracy 2.Use radioactive tracers to test downward transport from the stratosphere, including 7 Be 3.Global model for OH – test with MCF  Use observed distributions of CH 4, CO, O 3, NO x, and H 2 O  Compute the time evolution of MCF 4.Global model for CO with sources prescribed 5.Tropospheric ozone as an active tracer

The holy grail – tropospheric ozone “A reliable description of ozone presumes a model for NO x, for CO, for H 2 O, for OH and for heterogeneous chemistry, in addition to a satisfactory representation of the stratosphere and a valid description of troposphere-stratosphere exchange. It is unlikely that we can complete work on such a model within three years, but we expect to make substantial progress.” MBM, 1981.

Wet and dry convection Major steps in development of the chemical tracer model (CTM)  1986: Prather – 2 nd order moments scheme for accurate non-diffusive 3-d advection  1987: Prather et al. – CFCs as Tracers of Air Motion development of the CTM, many technical issues described sub-grid diffusion needed for correct interhemispheric gradient Higher resolution window 1985: Another new post-doc

 1990: Spivakovsky, Wofsy and Prather – chemistry parameterization for computation of OH  1990: Spivakovsky et al. Tropospheric OH in a 3-d CTM – an assessment based on MCF used observations of CO, CH 4, O 3, NO t, and column O 3 MCF lifetime with model OH is 5.5 y, MCF data implies 6.2 y OH fields provided to the community Updated and extended in Spivakovsky et al. (2000)  1990: Jacob and Prather – 222 Rn as a test of convective transport in a GCM

And finally, ozone, 12 years after the original proposal  1993: Jacob et al. – summertime ozone over the US used 6 tracers, parameterized chemistry sub-grid power-plant and urban plumes (Sillman et al., 1990) observations for boundary conditions gridded emissions from EPA, isoprene emissions dry deposition, wet deposition (Balkanski et al. 1993) evaluated with observations

And 17 years after MBM’s proposal – global ozone  1998: Wang, Jacob, and Logan – 3 papers on global model of O 3 -NO x -hydrocarbons 15 chemical tracers, new parameterizations global emission inventories from fossil fuel/industry biomass burning inventory, biogenic emissions lightning NOx stratospheric ozone flux extensive evaluation with observations All of the above work used the GISS GCM fields  2001: Bey et al. – Global tropospheric chemistry with assimilated meteorology – GEOS-Chem model GEOS met. fields from NASA Gear solver for chemistry (in window model earlier) Adopted many features from GCM based CTM

Déjà vu: from a “window” in 1987 to a nested grid formulation  2004: Yuxuan Wang, McElroy et al. – nested grid model for Asia. 1° x 1° resolution CO data model,1° x 1° Model, 4° x 5° Aircraft observations downwind of Asia in 2001 Applications to CO, NO x, ozone – tomorrow, China Project talks

Methane hindcast with a CTM,  2004: J. Wang, Logan, McElroy et al. – causes of the slowdown and variability in the CH 4 growth rate Data Slowdown in growth rate: slower growth in sources - the economic downturn in the former Eastern bloc increases in OH - column ozone decr. (solar cycle + trends) Variability wetland emissions + OH (especially post-Pinatubo) Results also showed model OH is too high (Wang et al., 2008) CH 4 growth rate, ppb/yr

Tropospheric chemistry in the 21 st century The NASA “A-Train” Surface sites Models Satellites Aircraft, ships, sondes, lidars Satellites provide a global continuous mapping of atmospheric composition, augmenting the otherwise sparse observing system Terra – CO data since 2000; Aura – CO, O 3, NO 2, HCHO since 2004 D.J. Jacob

NITROGEN DIOXIDE POLLUTION MEASURED FROM SPACE BY OMI 13 x 24 km pixels USED to CONSTRAIN SOURCES March 2006

MAPPING OF REACTIVE HYDROCARBON EMISSIONS FROM SPACE using measurements of formaldehyde columns Millet et al. [2008] Biogenic isoprene is the main reactive hydrocarbon precursor of ozone …and a major source of organic particles hydro- carbons 340 nm formaldehyde Jacob slide

Model inversion of CO sources using data from three satellite instruments – Kopacz et al., 2010 AIRS TES SCIAMACHY MOPITT Annual emissions Correction factors from inversion Errors of up to a factor of 2!

CO data for the upper and lower troposphere: a test of model transport MLS data GEOS-4 model CO (ppb) at 200 hPa Oct Nov  Fires in Aug/Sept. are a large source of CO satellite CO data show timing is correct in the lower troposphere  Convection moves south in October, lifting CO to the UT model peak is 1-3 months too late (GEOS-5 is worse than GEOS-4)  Detailed analysis shows: convection over South America detrains at too low an altitude too strong export of CO to the eastern Pacific in Aug./Sept isoprene is too large a source of CO  These transport problems will impact inversion studies, which cannot account for systematic errors in transport J. Liu, draft paper

Concluding remarks  CTM studies often identify significant problems with model transport – but they don’t necessarily lead to improvements in parameterizations inherent in global GCMs – an issue for 20+ years – 2 communities  Need for a holistic approach with satellite data – a tendency for one species per paper, and global data for CO, O 3, NO 2, HCHO, aerosols now available (also CH 4 )  CTMs now used for policy – e.g., long-range transport from Asia to the US, the US to Europe  Coupled chemistry-climate models used for projections  The stakes high – we would like to get things right!