Embed Size (px)
DESCRIPTION
Isoprene emissions in Africa inferred from OMI HCHO. Eloïse Marais ([email protected]) 1 , D. Jacob 1 , T. Kurosu 2 , K. Chance 3 , J. Murphy 4 , C. Reeves 5 , G. Mills 5 , S. Casadio 6 , D. Millet 7 , M. Barkley 8 , F. Paulot 1 , J. Mao 9 - PowerPoint PPT Presentation
Citation preview
Isoprene emissions in Africa inferred from OMI HCHO
ACKNOWLEDGEMENTS: This work was funded by the NASA ACMAP program and the South African National Research Foundation
OMI slant column (SC) HCHO (original data for 2005-2009 at 11°)
OMI SC HCHO(excludes biomass burning)
OMI SC HCHO(biogenic only)
OMI vertical column (VC) HCHO (biogenic only)
OMI VC HCHO(excludes smearing)
OMI-derived EISOP
(Error of ~40-90%)
1
MODIS fire counts remove HCHO at the source of burning.
OMI smoke AAOD tracks HCHO produced in biomass burning plumes transported extensively across Africa. Regional thresholds, used to identify biomass burning plumes, are cut-off values between biomass- and non-biomass- burning populations.MODIS fire counts
OMI smoke AAOD regional thresholds
HCHO produced during flaring of natural gas
(predominantly in Nigeria, Libya, Algeria, and Egypt)
filtered using AATSR hotspots
AATSR Algorithm 3 hotspots in 2005
100s of kilometers
Air mass factor (AMF) for 2005-2009
OMI surface albedo at 345 nm
R2 = 0.62Advection(low OH)OH, O2
NO
2
34
5
Displacement of HCHO from the isoprene emission source leads to smearing
MEGAN EISOP
Eloïse Marais ([email protected])1, D. Jacob1, T. Kurosu2, K. Chance3, J. Murphy4, C. Reeves5, G. Mills5, S. Casadio6, D. Millet7, M. Barkley8, F. Paulot1, J. Mao9
1Harvard University, 2JPL, 3Harvard-Smithsonian, CfA, 4U. Toronto, 5U. East Anglia, 6IDEAS, 7U. Minnesota, 8U. Leicester, 9U. Princeton
HCHO yield (S = HCHO/EISOP) vs column NO2 (NO2) from GEOS-Chem
STEPS 1-3:Isolate biogenic signal & convert HCHO SC to VC
STEP 4:Remove grid squares
influenced by smearing
STEP 5:Calculate EISOP with GEOS-Chem, MEGAN & OMI NO2
HighNOx
Low OH; high EISOP
~200-300 km smearing to the west in central Africa (region with high EISOP and low levels of OH)
Smearing is < 100 km over the AMMA domain (relatively high NOx conditions)
Surface albedo explains 62% of AMF variability
GLC2000 landcover at 0.250.25°
OMI biogenic HCHO is spatially consistent with
biome distribution in Africa
HCHO yields as a function of NO2 from GEOS-Chem are
applied to OMI NO2 to generate HCHO yields at 11°
OMI NO2
(no biomass burning)
Grid squares impacted by displacement of HCHO (and filtered out using GEOS-Chem) include grid squares located west of the Congo Basin and with low EISOP .
Monthly-average yields of HCHO at 11° are applied to OMI VC HCHO to determine OMI-derived EISOP
OMI-MEGAN
HCHO yields [103 s]
discrepancies
60 Tg C y-1
77 Tg C y-1
OMI is 33% lower than MEGAN in the tropics
OMI is 38% higher than MEGAN for closed and open broadleaf trees in s. Africa
Isoprene accounts for ~50% of global emissions of non-methane volatile organic compounds. It is a major precursor of secondary organic aerosols and tropospheric ozone, impacting human health, climate and air quality. Formaldehyde satellite observations (HCHO) have been used to better quantify isoprene emissions (EISOP) and test current emission inventories, such as the Model of Emissions of Gases and Aerosols from Nature (MEGAN). Africa is a major source of E ISOP, but has thus far received little attention.
Here we estimate EISOP for Africa using OMI HCHO observations during 2005-2009. We remove HCHO produced from biomass burning and anthropogenic sources with proxy observations from space. We identify and exclude HCHO displaced from the source of EISOP with the GEOS-Chem chemical transport model and MEGAN.
Procedure to isolate a biogenic HCHO signal from OMI and convert biogenic HCHO to EISOP
Colour Scale Key:OMI HCHO:
OMI-MEGAN:
EISOP:
Smearing under low NOx conditions makes the largest contribution (35-84%) to errors in the OMI-derived isoprene emissions.
