MAPPING OF VOLATILE ORGANIC COMPOUND (VOC) EMISSIONS USING SATELLITE OBSERVATIONS OF FORMALDEHYDE COLUMNS Daniel J. Jacob with Paui I. Palmer, Tzung-May Fu, Dylan B. Millet, Dorian S. Abbot and Kelly V. Chance, Thomas Kurosu (Harvard SAO/CFA) supported by NASA Atmospheric Chemistry Modeling and Analysis Program

SATELLITE OBSERVATIONS OF TROPOSPHERIC COMPOSITION …a rapidly growing resource! Sensor TOMS GOME IMG MOPITT MISR MODIS AIRS SCIA-MACHY TES OMI Platform (launch) multi (1979-) ERS-2 (1995) ADEOS (1996) Terra Aqua (1999) (2002) Envisat (2002) Aura (2004) ozone X (tropics) X aerosol CO NO2 HNO3 HCHO SO2 BrO

IMPORTANCE OF NON-METHANE VOC EMISSIONS FOR ATMOSPHERIC CHEMISTRY Precursors of tropospheric ozone Precursors of organic aerosols Sinks of OH Isoprene, terpenes, oxygenates… Alkenes, aromatics, oxygenates… Alkanes, alkenes, aromatics… ~ 200 Tg C yr-1 ~ 600 Tg C yr-1 ~ 50 Tg C yr-1 Vegetation Biomass burning Industry

SPACE-BASED MEASUREMENTS OF HCHO COLUMNS AS CONSTRAINTS ON VOC EMISSIONS solar backscatter 340 nm hn (l < 345 nm), OH Oxidation (OH, O3, NO3) VOC HCHO lifetime of hours many steps Emissions

SPACE-BASED MEASUREMENTS OF ATMOSPHERIC COLUMNS BY SOLAR BACKSCATTER Examples: TOMS, GOME, SCIAMACHY, MODIS, MISR, OMI, OCO Applications to retrievals of O3, NO2, HCHO, BrO, CO, CO2, aerosols… absorption Backscattered intensity IB l1 l2 wavelength Slant optical depth Scattering by Earth surface and by atmosphere “Slant column” Vertical column The air mass factor (AMF) depends on viewing geometry and radiative transfer

THE GOME INSTRUMENT Instrument in polar sun-synchronous orbit, 10:30 a.m. observation time 320x40 km2 field of view, three cross-track scenes Complete global coverage in 3 days Operational since 1995 HCHO column is determined from backscattered solar radiance in 340 nm absorption band Concurrent retrievals of cloud fractions, tops, optical depths

FITTING OF HCHO SLANT COLUMNS FROM GOME SPECTRA ts = 1.0 ± 0.3 x1016 cm-2 Fitting uncertainty of 4x1015 molecules cm-3 corresponds to ~ 1 ppbv HCHO in lowest 2 km ts = 3.0 ± 0.4 x1016 cm-2 ts = 8.4 ± 0.7 x1016 cm-2 Chance et al. [2000]

HCHO SLANT COLUMNS MEASURED BY GOME (JULY 1996) 2.5x1016 molecules cm-2 2 1.5 1 detection limit 0.5 South Atlantic Anomaly (disregard) -0.5 High HCHO regions reflect VOC emissions from fires, biosphere, human activity

AIR MASS FACTOR (AMF) CONVERTS SLANT COLUMN WS TO VERTICAL COLUMN W “Geometric AMF” (AMFG) for non-scattering atmosphere: q EARTH SURFACE

IN SCATTERING ATMOSPHERE, AMF DEPENDS ON VERTICAL DISTRIBUTION OF ABSORBER 340 nm HCHO EARTH SURFACE Use GEOS-Chem chemical transport model to specify shape of vertical profile for given scene

AMF FOR A SCATTERING ATMOSPHERE AMFG = 2.08 actual AMF = 0.71 GOME sensitivity w(z) HCHO mixing ratio profile S(z) (GEOS-Chem) what sees Palmer et al. [2001]

QUANTIFYING AMF ERRORS USING AIRCRAFT PROFILES ICARTT mission over North America (summer 2004) 0-10 km spirals and profiles during ICARTT: In situ HCHO, clouds, aerosol extinction Mean HCHO profiles in ICARTT Observed (Fried) Observed (Heikes) GEOS-Chem model (n = 89) Dylan B. Millet, Harvard Clouds are the principal source of error

FORMALDEHYDE COLUMNS FROM GOME: July 1996 means …compare to GEOS-Chem including GEIA biogenic VOC emissions and U.S. EPA anthropogenic VOC emissions GEOS-Chem vs. GOME: R = 0.83, bias = +14% Palmer et al. [2003]

SEASONALITY OF GOME HCHO COLUMNS (9/96-8/97) largely reflects seasonality of isoprene emissions GOME GEOS-Chem (GEIA) GOME GEOS-Chem (GEIA) MAR JUL APR AUG SEP MAY JUN OCT Abbot et al. [2003]

INTERANNUAL VARIABILITY OF GOME HCHO COLUMNS Augusts 1995-2001: correlation with temperature anomaly explains some but not all of the HCHO column variability GOME HCHO Temp. anomaly GOME HCHO Temp. anomaly 1995 1999 1996 2000 2001 1997 Abbot et al. [2003] 1998

RELATING HCHO COLUMNS TO VOC EMISSION oxn. hn (<345 nm), OH VOCi HCHO yield yi k ~ 0.5 h-1 Emission Ei smearing, displacement In absence of horizontal wind, mass balance for HCHO column WHCHO: Local linear relationship between HCHO and E … but wind smears this local relationship between WHCHO and Ei depending on the lifetime of the parent VOC with respect to HCHO production: Isoprene WHCHO a-pinene propane 100 km Distance downwind VOC source

TIME-DEPENDENT HCHO YIELDS FROM VOC OXIDATION Box model simulations with state-of-science MCM v3.1 mechanism methylbutenol High HCHO signal from isoprene with little smearing, weak and smeared signal from terpenes; GEOS-Chem yields may be too low by 10-40% depending on NOx Palmer et al, [2005]

HCHO COLUMN vs. ISOPRENE EMISSION RELATIONSHIP IN GEOS-Chem MODEL Results for U.S. quadrants in July 1996 simulation w/ 2ox2.5o horizontal resolution show: (1) dominance of isoprene emission as predictor of WHCHO variability; (2) linear relationship between the two Standard simulation NW NE R2 = 0.43 HCHO from simulation w/o Isoprene emission R2 = 0.51 Model HCHO column [1016 molec cm-2] R2 = 0.65 SW SE R2 = 0.49 We use this relationship to derive “top-down” isoprene emissions from the GOME HCHO column observations Isoprene emission [1013 atomC cm-2 s-1]

GOME vs. MEGAN ISOPRENE EMISSION INVENTORIES (2001) MEGAN is a new inventory of isoprene emissions developed by Alex Guenther [Guenther et al., 2005] Good accord for seasonal variation, regional distribution of emissions; GOME 10-34% higher than MEGAN depending on month, differences in hot spot locations Palmer et al. [2005]

EVALUATING GOME ISOPRENE EMISSION ESTIMATES vs EVALUATING GOME ISOPRENE EMISSION ESTIMATES vs. IN SITU FLUX MEASUREMENTS (2001) PROPHET forest site in northern Michigan (M. Pressley, WSU): also shown are local MEGAN isoperene emission inventory values Palmer et al. [2005]

YEAR-TO-YEAR VARIABILITY OF GOME HCHO OVER SOUTHEAST U.S. Amplitude and phase are highly reproducible GOME HCHO Column [1016 molec cm-2] Southeast US average 32-38N; 100-85W P. I. Palmer (Harvard)

WHAT DRIVES GOME HCHO TEMPORAL VARIABILITY OVER SOUTHEAST U. S WHAT DRIVES GOME HCHO TEMPORAL VARIABILITY OVER SOUTHEAST U.S. DURING MAY-SEPTEMBER? Monthly mean GOME HCHO vs. surface air temperature; MEGAN parameterization shown as fitted curve Temperature drives ~80% of the variance of monthly mean HCHO columns P.I. Palmer (Harvard)

GOME HCHO COLUMNS OVER EAST ASIA (1996-2001) JAN APR JUL OCT FEB MAY AUG NOV MAR JUN SEP DEC Relationship to VOC emissions far more complex than for N. America; biomass burning, isoprene, anthropogenic VOCs, direct HCHO emission all contribute Tzung-May Fu (Harvard)

GOME vs. GEOS-Chem HCHO COLUMNS OVER EAST ASIA MEGAN biogenic emission inventory is far too low APR MAY JUN JUL AUG SEP T. M. Fu (Harvard)

VOC CONTRIBUTIONS TO HCHO PRODUCTION IN CHINESE CITIES (JAN-FEB 2001) Vehicle-generated xylenes could make a large contribution to HCHO columns NC CC Ethane 0.3 % Benzene 0.4 % Propane 0.3 % Toluene 2.4 % ALK4 5.1 % Xylene 20.2 % Ethene 19 % Isoprene 8.2 % PRPE 43 % WC SC B. Barletta (UCI), T.-M. Fu (Harvard)

PRELIMINARY HCHO COLUMN DATA FROM OMI (launched on Aura in July 2004) 26 Day Average: 24 September – 19 October 2004 K. Chance and T. Kurosu (Harvard CFA)

OMI HCHO RETRIEVALS AND MODIS FIRE COUNTS Chongqing (Red Basin) Jakarta