Constraining Emissions with Satellite Observations Daniel Jacob Paul

Constraining Emissions with Satellite Observations Daniel Jacob Paul Palmer Dorian Abbot Randall Martin Kelly Chance Thomas Kurosu Sushil Chandra Jerry Ziemke

PRESENT AND FUTURE SATELLITE OBSERVATIONS OF TROPOSPHERIC CHEMISTRY 2002 2004 HCHO 2002 2004

CHALLENGE FOR THE NEXT DECADE: IMPROVE EMISSION INVENTORIES “Nova Scotians with lung problems should stay inside until pollution levels drop warned Environment Canada”

Historical records imply a large anthropogenic contribution to the present-day ozone background at northern midlatitudes Ozone trend from European mountain observations, 1870 -1990 [Marenco et al. , 1994] Major greenhouse gas Largely controls atmospheric oxidation Primary constituent of smog

GLOBAL BUDGET OF TROPOSPHERIC OZONE Global sources and sinks, Tg O 3 yr-1 (GEOS-CHEM model) Chem prod in troposphere 4900 Chem loss in troposphere 4200 Transport from stratosphere 500 Deposition 1200 ? Formaldehyde (HCHO) ? ? Nitrogen oxides (NOx) CO, VOCs Fires hv Biosphere Human activity Ozone (O 3) hv, H 2 O Hydroxyl (OH)

BOTTOM-UP EMISSION INVENTORIES ARE NOTORIOUSLY DIFFICULT TO DETERMINE! • • • Fuel use estimates Measurements of emission ratios Process studies Estimate biological density Temperature, water, … dependence of biological activity • Extreme events • "Trees cause more pollution than automobiles do. "

HOW DO WE EVALUATE AND IMPROVE A PRIORI BOTTOMUP INVENTORIES? Surface NOX Isoprene during July GEIA

TOP-DOWN INFORMATION FROM THE GOME SATELLITE INSTRUMENT • Operational since 1995 • Nadir-viewing solar backscatter instrument (237 -794 nm) • Low-elevation polar sunsynchronous orbit, 10: 30 a. m. observation time • Spatial resolution 320 x 40 km 2, three cross-track scenes • Complete global coverage in 3 days

USE GOME MEASUREMENTS TO RETRIEVE NO 2 AND HCHO COLUMNS TO MAP NOx AND VOC EMISSIONS GOME Tropospheric NO 2 column ~ ENOx Tropospheric HCHO column ~ EVOC BOUNDARY LAYER NO/ NO 2 NO NO 2 hours W ALTITUDE lifetime <1 day CO HCHO hours OH VOC HNO 3 Emission NITROGEN OXIDES (NOx) Emission VOLATILE ORGANIC COMPOUND (VOC)

PERFORM A SPECTRAL FIT OF SOLAR BACKSCATTER OBSERVATIONS absorption Solar Io Backscattered intensity IB l 1 l 2 wavelength Slant optical depth Scattering by Earth surface and by atmosphere “Slant column” EARTH SURFACE

GOME HCHO SLANT COLUMNS (JULY 1996) Hot spots reflect high VOC emissions from fires and biosphere BIOGENIC ISOPRENE IS THE MAIN SOURCE OF HCHO IN U. S. IN SUMMER K. Chance

SLANT COLUMNS OF NO 2 FROM GOME Dominant stratospheric structure (where NO 2 is produced from N 2 O oxidation) Also see tropospheric hot spots (fossil fuel and biomass burning) Remove strat & instrument artifacts using obs over Pacific

SLANT COLUMNS OF TROPOSPHERIC NO 2 FROM GOME 1996

GEOS-CHEM MODEL • • • Assimilated Meteorology (GEOS) 2 ox 2. 5 o horizontal resolution, 26 layers in vertical O 3 -NOx-hydrocarbon chemistry Radiative and chemical effects of aerosols Emissions: – Fossil fuel: GEIA (NOx), Logan (CO), Piccot (NMHCs) – Biosphere: modified GEIA (hydrocarbons) & Yienger/Levy (soil NOx) – Lightning: Price/Rind/Pickering, GEOS convective cloud tops – Interannually varying biomass burning • Cross-tropopause transport • Deposition

IN SCATTERING ATMOSPHERE, AMF CALCULATION NEEDS EXTERNAL INFO ON SHAPE OF VERTICAL PROFILE RADIATIVE TRANSFER MODEL “a priori” Shape factor Io sigma ( ) IB ATMOSPHERIC CHEMISTRY MODEL dt( ) EARTH SURFACE Scattering weight ( ) is temperature dependent cross-section Calculate w( ) as function of: • solar and viewing zenith angle • surface albedo, pressure • cloud optical depth, pressure, frac • aerosol profile, type NO 2 mixing ratio CNO 2( ) norm. by column ΩNO 2 INDIVIDUAL GOME SCENES

VERTICAL COLUMNS CONFINED TO REGIONS OF SURFACE EMISSIONS Cloud artifacts removed by AMF calculation NO/ NO 2 WITH ALTITUDE NOx lifetime <1 day

GOME Tropospheric NO 2 GEOS-CHEM Tropospheric NO 2 r=0. 75 bias 5% 1015 molecules cm-2

STRATEGY: OPTIMIZE INVENTORIES USING A PRIORI BOTTOM-UP AND GOME TOP-DOWN INFORMATION Retrieved emissions A priori emissions A posteriori emissions Retrieval errors A priori errors

TOP-DOWN ERROR IN NOX EMISSIONS GOME Spectrum (423 -451 nm) Spectral fit and removal of stratospheric column 1 x 1015 molecules cm-2 Tropospheric NO 2 Slant Column AMF (surface reflectivity, clouds, 42% of tropospheric column aerosols, NO 2 profile) Tropospheric NO 2 Column NOx Lifetime (GEOS-CHEM) 30% of tropospheric column NOx Emissions

TOP-DOWN INFORMATION FROM GOME REDUCES ERROR IN NOX EMISSION INVENTORY Bottom-up error a Top-down error r

OPTIMIZED NOX EMISSIONS

DIFFERENCE BETWEEN A POSTERIORI AND A PRIORI Annual mean ratio (A posteriori / A priori)

TROPOSPHERIC OZONE DETERMINED FROM TOMS/MLS SUGGESTS A PRIORI NOX EMISSIONS HIGH OVER INDIA Dobson Units for March, April, May TOMS/MLS Chandra et al. [2003] GEOS-CHEM

ISOPRENE EMISSIONS FOR JULY 1996 DETERMINED BY SCALING GOME FORMALDEHYDE COLUMNS GOME COMPARE TO… GEIA (IGAC inventory) [Palmer et al. , 2003]

EVALUATE GOME ISOPRENE INVENTORY BY COMPARISON WITH IN SITU OBSERVATIONS USING GEOS-CHEM MODEL AS INTERMEDIARY GEIA (A priori) r 2 = 0. 53 CONSISTENT WITH IN-SITU HCHO OBSERVATIONS GOME (A posteriori) r 2 = 0. 77 Palmer et al. , 2003