Modelling Biomass Burning Emissions and their Optical Properties

1
Modelling Biomass BurningEmissions and their
Optical Properties
School of Earth Environment Institute for
Atmospheric Science

• David Ridley
• Ken Carslaw, Martyn Chipperfield, Graham Mann

2
What is Biomass Burning?

• Burning of vegetation for fuel, to clear land,
  remove crop debris, or to perform rituals.
• Estimated to be 90 anthropogenic
• 32 - Forests (20 of this is replanted)
• 60 - Grasslands
• 8 - Croplands

3
Climatic Effects

• Warming due to release of greenhouse gases

• Biomass burning releases gases (e.g., CO2, CO,
  CH4, NOx, SO2, C2H6, C2H4, C3H8, C3H6)
• A third of all anthropogenic CO2 emissions are
  from biomass burning
• Totally global forcing of roughly 0.5Wm-2

4
Climatic Effects

• Cooling due to release of aerosol - Black Carbon
  (BC), Particulate Organic Matter (POM)

• Increase in aerosol reduces light reaching the
  surface (Direct Effect)
• -0.1 to -0.6Wm-2 global forcing due to biomass
  burning (IPCC, 2001)
• Increased aerosol can increase cloud brightness
  and reduce rainfall (Indirect Effects)
• -0.1 to -0.4Wm-2 forcing due to biomass burning
  globally (Jacobson, 2004)
• Black carbon particles can have a warming effect
  upon atmosphere (Semi-Direct Effect)
• ??Wm-2 to ??Wm-2 very hard to constrain

5
Locations

• Primarily occurs in South America, Africa and
  Indonesia

J. Heintzenberg
Direct Radiative Forcing
6
UKCA Optical Depth

• Model currently under-predicts AOD
• Why?

MODIS AOD Oct 2000
UKCA AOD Oct 2000
7
Modelling Optical Depth

• Factors affecting reproduction of observed AOD
  over source
• Mass of Emissions
• Injection Height
• Emission Size Distribution
• Aerosol Aging
• Wet Deposition
• Spatial Temporal Variability

8
Determining Emissions

• Location and Properties of Fire Emissions
• Identifying location and duration of fire
• Satellite detection of fires and burnt areas
• Estimating amount and biome type of the burnt
  biomass (available fuel load)
• Climatological Biomass Maps and NDVI from
  satellite
• Choosing the right specific species emission
  factors and injection height for the detected
  fire
• In-situ concentration measurements (Andreae and
  Merlet, 2001)

(Liousse, C. et al., 2004).
9
Emissions Databases

• GFED
• 1x1deg, Monthly and Climatological Averages
• Uses ATSR fire detection
• Emission heights from Olsen vegetation map
  (Lavoue thesis, 2004)
• RETRO
• 0.5x0.5deg, monthly average
• Uses MODIS fire detection
• No emission heights

10
Emissions Database Comparison

• Significant variability in location and magnitude
  of emissions

Testing of the effect of different databases and
injection heights upon AOD and aerosol size
distribution to be conducted
11
Injection Size

• Model currently assumes injection size of 75nm
  radius

2nd biomass mode
Observations (Haywood, 2003) suggest 100nm may be
more appropriate for a modal scheme
12
Emission Size Distribution

• Comparison with AERONET size distributions show
  lack of AOD is primarily due to accumulation mode
  aerosol

70nm
100nm

• Increasing BC/OC effective radius by x1.75 and
  number by x3 required
• Suggests large under-estimation of mass! (or
  water uptake)

13
Emission Size Distribution

• Knowing mass of aerosol is not enough to
  determine AOD
• Size of aerosol has substantial effect upon AOD
  even when mass is held constant

14
Modelling Optical Depth

• Factors affecting reproduction of observed AOD
  over source
• Mass of Emissions
• Injection Height
• Emission Size Distribution
• Aerosol Aging
• Wet Deposition
• Spatial Temporal Variability

15
Aerosol Aging

• Water uptake important in AOD
• BC OC initially insoluble but become more
  soluble over time
• Altering solubility within reasonable limits
  give large change in AOD
• No direct effect upon number concentration
  only size

0 Soluble
10 Soluble
20 Soluble
16
Wet Deposition I

• Too much aerosol may be rained-out in the model
• Model rainfall much more intense than observed,
  but spatial distribution OK over region of
  interest i.e. very little rain

FEWS Rainfall
Model Rainfall
17
Wet Deposition II

• Efficiency of nucleation scavenging reduced from
  99.9 to 50
• Change of around 0.05 in AOD not sufficient to
  explain discrepancy

99.9
50
18
Modelling Optical Depth

• Factors affecting reproduction of observed AOD
  over source
• Mass of Emissions
• Injection Height
• Emission Size Distribution
• Aerosol Aging
• Wet Deposition
• Spatial Temporal Variability

19
Summary

• Several important factors influencing biomass
  AOD in model
• Discrepancies in spatial distribution of
  aerosol
• Likely to be due to emission number conc. rather
  than wet dep.
• Effect of injection height to be investigated
  may affect aerosol number concentration
• Will be quantified using different emissions
  databases
• Size of emissions important!
• Can cause doubling of AOD in heavily polluted
  regions (even when mass is held constant)
• Will be constrained using in-situ data to
  initiate emissions sizes
• Hygroscopicity important
• Crude test shows assumptions of BC/OC aging
  affect AOD by 25-75 (depending on initial size)