Ozone Monitoring Experiment Using space-based measurements of formaldehyde to learn about BVOC distributions Paul Palmer, University of Edinburgh biogenic, pryogenic, anthropogenic pryogenic anthropogenic biogenic pyrogenic Thomas Kurosu, Harvard-Smithsonian HCHO August 2006 Ozone Monitoring Experiment

Tropospheric O3 is an important climate forcing agent NO HO2 OH NO2 O3 hv HC+OH  HCHO + products NOx, HC, CO Level of Scientific Understanding Natural VOC emissions (50% isoprene) ~ CH4 emissions. IPCC, 2001

MEGAN Isoprene Emission Inventory Environmental factors: temperature solar irradiance leaf area index leaf age July 2003

Global Ozone Monitoring Experiment (GOME) & the Ozone Monitoring Instrument (OMI) Launched in 2004 GOME (European), OMI (Finnish/USA) are nadir SBUV instruments Ground pixel (nadir): 320 x 40 km2 (GOME), 13 x 24 km2 (OMI) 10.30 desc (GOME), 13.45 asc (OMI) cross-equator time GOME: 3 viewing angles  global coverage within 3 days OMI: 60 across-track pixels  daily global coverage O3, NO2, BrO, OClO, SO2, HCHO, H2O, cloud properties

HCHO Column Abundance Fitted in a Narrow UV Spectral Window 226-356 nm fitting window Fitting uncertainty of slant columns is typically < 4x1015 molec cm-2

vertical column = slant column /AMF GEOS-CHEM satellite lnIB/ Sigma coordinate () dHCHO 1 Earth Surface HCHO mixing ratio C() Scattering weights Shape factor S() = C() air/HCHO w() = - 1/AMFG lnIB/ 1 AMF = AMFG  w() S() d

GOME HCHO columns July 2001 Biogenic emissions Biomass burning – fractionally cloudy pixels (>40%) removed Biogenic emissions Biomass burning July 2001 Data: c/o Chance et al [1016 molec cm-2] 1 2 0.5 1.5 2.5 * Columns fitted: 337-356nm * Fitting uncertainty < continental signals

Relating HCHO Columns to VOC Emissions hours OH h, OH VOC Net kHCHO EVOC = (kVOCYVOCHCHO) HCHO ___________ Local linear relationship between HCHO and E Three/four prong attack: Chemistry; Emissions: magnitude and controls; Transport VOC source Distance downwind WHCHO Isoprene a-pinene propane 100 km EVOC: HCHO from GEOS-CHEM CTM and MEGAN isoprene emission model Palmer et al, JGR, 2003.

chemistry transport model Modeling Overview GEOS-CHEM global 3D chemistry transport model MCM: parameterized HCHO source from monoterpenes and MBO PAR, T Emissions MODEL BIOSPHERE MEGAN (isoprene) Canopy model Leaf age LAI Temperature Fixed Base factors GEIA Monoterpenes MBO Acetone Methanol Monthly mean AVHRR LAI

MCM HCHO yield calculations 0.5 NOx = 1 ppb NOx = 0.1 ppb Isoprene C5H8+OH(i) RO2+NOHCHO, MVK, MACR (ii) RO2+HO2ROOH ROOH recycle RO and RO2 Cumulative HCHO yield [per C] HOURS Higher CH3COCH3 yield from monoterpene oxidation  delayed (and smeared) HCHO production 0.4 Explain reasons for delayed production of HCHO from pinenes Parameterization (1ST-order decay) of HCHO production from monoterpenes in global 3-D CTM  pinene ( pinene similar) DAYS Palmer et al, JGR, 2006.

Seasonal Variation of Y2001 Isoprene Emissions MEGAN GOME MEGAN GOME May Jun Aug Sep Jul 3.5 7 1012 atom C cm-2s-1 Good accord for seasonal variation, regional distribution of emissions (differences in hot spot locations – implications for O3 prod/loss). Other biogenic VOCs play a small role in GOME interpretation Palmer et al, JGR, 2006.

Isoprene flux [1012 C cm-2 s-1] Sparse ground-truthing of GOME HCHO columns and derived isoprene flux estimates Isoprene flux [1012 C cm-2 s-1] Julian Day, 2001 MEGAN Obs GOME May Jun July Aug Sep Seasonal Variation: Comparison with eddy correlation isoprene flux measurements (B. Lamb) is encouraging Atlanta, GA 1996 1997 1998 1999 2000 2001 PROPHET Forest Site, MI Atlanta, GA To evaluate the GOME interannual variability over the southeastern United States we used isoprene concentration data from four EPA Photochemical Assessment Monitoring Sites (PAMS, http://www.epa.gov/air/oaqps/pams/) located around Atlanta Georgia. Three of these sites are classed as surburban and one is considered rural. Instantaneous concentration measurements are taken every three hours using an automated GC with flame ionization detection. The uncertainty of these individual measurements is 30\% (Susan Zimmer-Dauphinee, EPA, personal communication, 2004). Instruments are calibrated daily using an isoprene standard. Measurements that do not agree with the standard to within a 30\% accuracy are discarded. As with the HCHO columns, there is a large degree of interannual variability in the observed seasonal cycle on the continental scale (not shown). \callout{Figure \ref{fig:pamsga}} shows that GOME HCHO column data captures 58\% of the temporal variability of the monthly mean isoprene concentrations at all Atlanta sites, after removing one anomalous measurement (880~ppbC, July 1996) and two outliers (June 1998 and August 1999). % The value for the intercept (0.2$\times$10$^{16}$molec~cm$^{-2}$) in Figure \ref{fig:pamsga}, corresponding to the HCHO column with no contribution from isoprene, is half the background HCHO column that is expected from from CH$_4$ oxidation. The two monthly mean outliers originate from the rural site that is sometimes influenced by local biogenic emissions from the Kudzu vine, a known strong emitter of isoprene (Susan Zimmer-Dauphinee, EPA, personal communication, 2004). The reason why these local biogenic emissions significantly influence only a few months is unknown. By comparing mean summertime values (June$-$August) we effectively test the ability of GOME to capture observed interannual variability in isoprene concentration. GOME summertime columns between 1996$-$2001 capture 92\% of the observed interannual variability. GOME HCHO [1016 molec cm-2] Interannual Variation: Correlate with EPA isoprene surface concentration data. Outliers due to local emissions. PAMS Isoprene, 10-12LT [ppbC]

GOME Isoprene Emissions: 1996-2001 May Jun Jul Aug Sep 1996 1997 1998 1999 2000 2001 Palmer et al, JGR, 2006. [1012 molecules cm-2s-1] 5 10

Surface temperature explains 80% of GOME-observed variation in HCHO NCEP Surface Temperature [K] GOME Isoprene Emissions [1012 atoms C cm-2s-1] G98 fitted to GOME data G98 Modeled curves Palmer et al, JGR, 2006. Time to revise model parameterizations of isoprene emissions?

Tropical ecosystems represent 75% of biogenic NMVOC emissions What drives observed variability of tropical BVOC emissions?

Significant pyrogenic HCHO source over tropics Monthly ATSR Firecounts GOME Good: Additional trace gas measurement of biomass burning; effect can be identified largely by firecounts (see below) Bad: Observed HCHO a mixture of biogenic and pyrogenic – difficult to separate without better temporal and spatial resolution Sep 1997 1997 1998 1999 2000 2001 X = Active Fire (ATSR) Monthly ATSR Firecounts Slant Column HCHO [1016 molec cm-2] Nov 1997 Day of Year

HCHO and Isoprene over the Amazon In situ isoprene 2002 Trostdorf et al, 2004 1997 1998 1999 2000 2001 GOME ATSR Firecounts used to remove HCHO from fires

Can isoprene explain the observed magnitude and variance of HCHO columns over the tropics? Africa Isoprene Limonene Beta-pinene [ppb] Time of Day Amazon emission rate (C) (µg g-1 h-1) PAR (µmol m-2 s-1) assimilation (C) (mg g-1 h -1) 1 2 3 4 5 6 limonene myrcene b-pinene a-pinene sabinene 500 1000 1500 00:00 06:00 12:00 18:00 local time [hh:mm] 10 20 30 40 temperature [°C] G93 for isop. [sum of monoterpenes] transpiration (mmol m-2 s-1) monoterpene emission of Apeiba tibourbou C/o J. Kesselmeier C/o J. Saxton A. Lewis

OMI gives a better chance of estimating African BVOCs May Jun Jul Aug Sep Oct OMI Data c/o Thomas Kurosu; horizontal resolution O(10x25 km2)

Isoprene concentrations during AMMA July 2006 measured by the Bae146 aircraft; MODIS tree cover overlaid [ppt] OMI ATSR Firecounts Jul Jul Isoprene data c/o Jim Hopkins and Ally Lewis, U. York

Compiled from UK ozone network data An increasing role for BVOCs in UK air quality? “Normal” airmass flow Stagnant airmass flow 200 400 600 800 1000 1200 1400 27-Jul 29-Jul 31-Jul 2-Aug 4-Aug 6-Aug 8-Aug 10-Aug 12-Aug 14-Aug 16-Aug 18-Aug 20-Aug 22-Aug 24-Aug 26-Aug 28-Aug 30-Aug 5 10 15 20 25 30 35 40 Temperature (C) Isoprene (ppt) Estimated up to 700 extra deaths attributable to air pollution (O3 and PM10) in UK during this period O3 > 100 ppb on 6 consecutive days 2pm, 6th Aug, 2003 Compiled from UK ozone network data Isoprene is normally 2-50 ppt No temperature dependence given!!!! Same latitude as hudson bay!! Isoprene c/o Ally Lewis

The European Heatwave of 2006 Jun Jul Aug Stewart et al, 2003 Isoprene Monoterpenes BVOC fluxes for a “hot, sunny” day No temperature dependence on isoprene emission….spruce plantation on the borders… Tropospheric O3 production results from NO + RO2  NO2 + RO The influence of different VOCs on this step can be calculated Much higher total rate of NO  NO2 conversion during the heatwave period (NB – the graphs have different scales) Isoprene contributes substantially to O3 production Given the short lifetime of isoprene, it must have been generated and reacted locally Satellite observations test bottom-up emission inventories used for air quality: an important step toward regional chemical weather forecasting

Final Comments Proper interpretation of HCHO requires an integrated approach, i.e., including surface data, lab data Interpreting space-based HCHO data is still in its infancy – new instruments bring better resolution but also new challenges With the frequency of European heatwaves projected to increase the role of BVOCs in future UK air quality must be better quantified