1. Dylan Millet Harvard University with D. Jacob (Harvard), D. Blake (UCI), T. Custer and J. Williams (MPI), J. de Gouw, C. Warneke, and J. Holloway (NOAA), T. Karl (NCAR), H. Singh (NASA), B. Sive (UNH) New Constraints on Terrestrial and Oceanic Sources of Atmospheric Methanol NASA Atmospheric Chemistry Program NOAA C&GC Postdoctoral Fellowship Program American Geophysical Union Fall Meeting 2007 Thanks to:

2. Methanol: The Most Abundant Non-Methane Organic Gas CH 3 OH Burden: ~4 Tg Lifetime: 5-10 d CH 4 Plant Decay Biomass Burning Atmospheric Production Dry Dep (Land) Wet Dep Oxidation by OH Source of CO, HCHO, O 3 Sink of OH Plant Growth Ocean Exchange Urban Emissions ?

3. Aircraft and Surface Measurements Used to Constrain Methanol Sources & Sinks New plot with all obs AIRCRAFT PEM-TB, INTEX-A/B, MILAGRO, ITCT-2K2/2K4, TOPSE, LBA/CLAIRE, TROFFEE, TEXAQS-II SURFACE OOMPH, NEAQS-2K2, Kinterbish, Tennessee, UMBS, Trinidad Head, Duke Forest, Chebogue Pt, Appledore Isl., Thompson Farm, Rondônia, Amazonas GEOS-Chem 3D model of atmospheric chemistry Interpret with:

4. Methanol: The Most Abundant Non-Methane Organic Gas CH 3 OH Burden: ~4 Tg Lifetime: 5-10 d CH 4 Plant Decay Biomass Burning Atmospheric Production Dry Dep (Land) Wet Dep Oxidation by OH Source of CO, HCHO, O 3 Plant Growth Ocean Exchange Urban Emissions

5. Ocean Mixed Layer (OML): Source + Sink for Atmospheric Methanol Previous work: · Assume constant OML undersaturation · OML a small net sink  Assumes air-sea exchange controls [CH 3 OH] OML

6. Ocean Mixed Layer (OML): Source + Sink for Atmospheric Methanol Recent OML Measurements imply a large methanol reservoir (20× that of the atmosphere) Biotic consumption  ~ 3 d [Heikes et al., 2002] CH 3 OH 120 ± 50 nM [Williams et al., 2004] 66 Tg  Short lifetime requires large OML source (~8E3 Tg/y)

7. Ocean Mixed Layer (OML): Source + Sink for Atmospheric Methanol Recent OML Measurements imply a large methanol reservoir (20× that of the atmosphere) 100 Tg/y OML ventilation weeks-months Biotic consumption  ~ 3 d [Heikes et al., 2002] CH 3 OH 120 ± 50 nM [Williams et al., 2004] 66 Tg Biological production  Short lifetime requires large OML source (~8E3 Tg/y)  Transfer from atmosphere insufficient to balance loss  Large in-situ biological source implied

8. Ocean Mixed Layer (OML): Source + Sink for Atmospheric Methanol Recent OML Measurements imply a large methanol reservoir (20× that of the atmosphere) 100 Tg/y OML ventilation weeks-months Biotic consumption  ~ 3 d [Heikes et al., 2002] CH 3 OH 120 ± 50 nM [Williams et al., 2004] 66 Tg Biological production  Short lifetime requires large OML source (~8E3 Tg/y)  Transfer from atmosphere insufficient to balance loss  Large in-situ biological source implied  Ocean emission, uptake: independent terms in atmospheric budget

9. Ocean Emission and Uptake of Atmospheric Methanol  Marine biosphere: large source of atmospheric methanol  Comparable to terrestrial biota Ocean Emission 85 Tg y -1 Ocean Uptake  Comparable to oxidation by OH Calculate ocean source & sink terms independently · On basis of measured OML concentrations 100 Tg y -1  =11 d Net Flux

10. New Air-Sea Flux Parameterization Generally Consistent with Atmospheric Observations Measured vs. modeled methanol concentrations over the S. Atlantic OOMPH 2007 Measured Modeled Methanol profiles over the Pacific

11. Methanol Emissions from the Terrestrial Biosphere Aircraft Measurements Reveal Overestimate of Plant Growth Source All plants make methanol · Produced during cell growth · Emitted from leaves ~ f(T, hν) · E = 0.11% × NPP [Galbally & Kirstine, 2002] Simulated summer methanol concentrations in surface air [ppb] Measured Modeled Vertical Profiles over N. America  Broad-scale inflow to W. US well simulated  2× BL overestimate during summer  Only explained by overestimate of plant growth source

12. Bias Correlates Spatially with Regions of High Broadleaf Tree & Crop Coverage ObservedModeled Boundary Layer Methanol Concentrations [ppb] Modeled - Measured Removal of bias requires:  4x reduction of broadleaf tree + crop emissions, or  2x reduction of emissions from all terrestrial plants MDVD2 vegetation coverage [Guenther et al., 2006]

13. Reduced Biogenic Source Yields Better Agreement over North America and Tropical South America Measured Base case  2× (all plants)  4× (bdlf trees + crops) Vertical Profiles over N. AmericaAmazon Boundary Layer Base case  2× all plants  4× bdlf trees, crops Measured  Both optimizations of comparable quality  Best estimate of global terrestrial biogenic source: 80 Tg/y (vs. 145 Tg/y base case)

14. Importance of Biogenic vs. Anthropogenic Sources Methanol strongly correlated with CO despite lack of large anthro. source Aircraft measurements over N. America during summer Model captures correlation, slope (with independent constraints on CO) Measured Base case  2× (all plants)  4× (bdlf trees + crops)

15. Updated Global Budget of Atmospheric Methanol SourcesSinks 10 8 molec/cm 2 /s 85 Tg/y 80 Tg/y 37 Tg/y23 Tg/y 12 Tg/y5 Tg/y 101 Tg/y 88 Tg/y 40 Tg/y13 Tg/y Atmospheric lifetime: 4.7 days