Title: Direct and indirect climatic effects of aerosols from biomass burning in Amazonia
1Direct and indirect climatic effects of aerosols
from biomass burning in Amazonia
- Yoram J. Kaufman Symposium On Aerosols, Clouds
and Climate - May 30, 31, and June 1, 2007 NASA/Goddard Space
Flight Center
Paulo Artaxo, J. Vanderlei Martins, Melina
Paixão, Karla Longo, Saulo Freitas, Aline
Procópio, Andrea Castanho, Marcia Yamasoe, Joel
Schafer, Brent Holben, Tom Eck, and many
others University of São Paulo, Brazil,
artaxo_at_if.usp.br
2- Paul Crutzen in 1979 hypothesized that biomass
burning could be important to TRACE GASES
globally. (Aerosols were not important at that
time). - Yoram shortly after that started the BASE-A
Experiment (Biomass Burning airborne and
spaceborne Experiment in Amazonia), and latter
the planning for the SCAR-B Experiment. - Biomass Burning is relevant not only to global
climate. In the Milagro Megacity Experiment,
aerosol source apportionment calculations shows
that biomass burning accounts for 10-50 of PM2.5
in downtown Mexico City. Similar values for
other mega cities in Asia and India are possible.
(Madronich AGU Acapulco last week) - In the Western United States, forest fires in the
last 10 years have been responsible for up to a 6
ppb increase in the ozone concentrations, that is
a large share of ozone atmospheric
concentrations. (AGU Acapulco last week)
3Deforestation and fire spots in Amazonia
4CO2 emission inventory for Brazil
Brazil is the fifth largest CO2 contributor in
1994, Without biomass burning, it would be the
16º.
74 of Brazilian emissions is related to
deforestation.
5- The most important air pollution issue in South
America is associated to the continental scale
biomass burning during the dry season. With
several hundred of thousands of fires each year - Severe health effects on the population
- Climate effects
- Weather effects
MODIS Rapid Response System 08/25/2004 1415UTC
6The Amazonian green ocean Oceanic clouds over
continents
7Pyrocumulus Clouds
Hydrological cycle critical for Amazonia. Variety
of cloud structure caused by different CCN
amounts and other cloud dynamic issues
Green Ocean Clouds
8Pyrocumulus Clouds
9Addition of pyrogenic CCN has pronounced impact
on cloud droplet size spectra
Four aerosol regimes of(A) Blue Ocean,(B) Green
Ocean,(C) Smoky clouds, (D) Pyro-clouds Note
that the narrowing of CDSD and the slowing of its
rate of broadening with height for the
progressively more aerosol rich regimes from A to
D.
D. Rosenfeld, 2004
10Aerosol-cloud interactions
11/August/2002 Brazilian Amazon
15/November/2002 Brazilian Amazon
Satellite images of the Amazon rainforest rarely
show smoke and low cumulus clouds together.
Satellite images of the Amazon rainforest rarely
show smoke and cumulus clouds together. Smoke,
mainly from agricultural fires, displaces the
cumulus clouds that normally form above the
forest each afternoon.
A uniform layer of scattered cumulus clouds is
typically present, along with some thunderstorms,
over the Amazon rainforest. Compare this image of
a day with little smoke, with the image above.
Both images were acquired by the Moderate.
Koren et al., 2004 - Science
11Large scale low cloud suppression
Terra and Aqua satellite images of the east
Amazon basin, 11 August 2002. (A) The clouds
(Terra, 1000 local time) are beginning to form.
(B) The clouds (Aqua, 1300 local time) are fully
developed and cover the whole Amazon forest
except for the smoke area. The boundary between
forest and Cerrado region is marked in white on
both images, and the seashore is marked in green.
(From Koren et al., 2004)
12Suppression of low cloud formation by aerosols in
Amazonia
Cloud fraction as function of aerosol optical
depth (OD). The cloud fraction decreases almost
linearly with increasing OD. The red and blue
curves denote the average of east and west areas,
respectively. On average, the cloud fraction
decreases to less than 1/8 of the cloud fraction
in clean conditions when OD 1. The shaded area
represents the relative area covered by the
respective OD, with the integral of this curve
equal to one, representing the total Amazon
basin. (from Ilan and Kaufman, 2003)
13Concentrações de partículas de aerossóis (PM10)
em Alta Floresta 1992-2000
14INDOEXaverage aerosol forcing clear sky
AmazoniaAverage aerosol forcing clear sky
Top - 71 w/m²
Top - 10 w/m²
Atmosphere 162 w/m²
Atmosphere 28 w/m²
Surface - 38 w/m²
Surface - 232 w/m²
Conditions surface ocean AOT (?0.3 at 630 nm)
24 hour averageJan-Mar 99
Conditions surface forest vegetation AOT
(?0.95 at 500nm) 24 hour average7 years
(93-95, 99-02 dry season Aug-Oct)
Procópio et al. (2004)
15Aerosol surface forcing in Rondonia 1999-2002
Aline Procópio
16The Large Scale Biosphere Atmosphere Experiment
in Amazonia - LBA
Aerosols(and trace gases)
Water (in clouds and biosphere)
Anthropogenic activities
Carbon (Vegetation and soil)
Nutrients (P, N, K, others)
17Effects of biomass burning aerosols on
photosynthetic rate of the Amazonian forest
CO2 Concentration
Aerosol Concentration
-
Temperature
?
Photosynthesis
BVOC emissions
Kulmala et al., 2004
18Relative irradiance versus AOT
CO2 flux versus AOT
?AOT 0.1 to 1.2 ? ? fB - 20 ? ? NEE 50
After doing cloud screening to get only the
aerosol effect, the result is shown in (a). The
irradiance is reduced up to 20 for AOT varying
from 0.1 to 1.2. The relationship between CO2
fluxes measurements (NEE, storage corrected) and
aerosols (b), expressed as AOT. The CO2 flux
increases up to 50 for AOT varying from 0.1 to
1.2. The increase in the diffuse fraction of the
solar irradiance is the factor that explain this
behavior.
19Forte efeito das partículas de aerossóis na
fotossíntese da floresta
20AERONET sites in operation in Amazonia
21Aeronet aerosol optical thickness for Rondonia
and Alta Floresta
22Weekly averages of aerosol optical depth (440nm)
for regionally-grouped sites
Schafer et al., 2007
23Aerosol volume size distributions averaged for
each aerosol optical depth interval
Cerrado site (Cuiabá)
Southern forest site (Alta Floresta)
Volume median radius of the fine mode (VMRf) and
fine mode fraction of AOD averaged for each AOD
bin are also shown.
Schafer et al., 2007
24Averaged volume size distributions (fine mode
only) grouped by aerosol optical depth for the
composite Amazonian southern forest sites and for
an urban aerosol (GSFC) for all cases when the
observed column water vapor was between 3.5 and
4.5 cm
Averaged volume size distributions (fine mode
only) grouped by aerosol optical depth for the
composite Amazonian southern forest sites
partitioned into two ranges of column water
vapor.
Schafer et al., 2007
25Trends of single scattering albedo (SSA550nm) as
a function of column water vapor for two aerosol
optical depth bins.
Average SSA Forest North 0.914 Forest South
0.926 Cerrado 0.887
Schafer et al., 2007
26Histograms of single-scattering albedo (SSA550nm)
measurements for the southern forest and cerrado
sites (all years combined).
Schafer et al., 2007
27Histograms of single-scattering albedo (SSA550nm)
measurements for Alta Floresta and Cuiabá (all
years combined) partitioned into two AOD groups
moderate (AOD 0.4 to 0.7) plots on the left and
high (AOD gt 1.0) plots on the right.
Schafer et al., 2007
28Isoprene
2-methilthertiol
From Clayes et al., Science March 2004)
- Natural Production of CCN in Amazonia
- Primary biogenic particles acting as giant CCN
- Secondary organic aerosol from terpenes, isoprene
and others - Coated soil dust (very small)
- Sulfates and nitrates (small)
29CCN sources in the clean tropics
- Dust, Sea Spray Over the humid continental
tropics, too few to be important - Marine biogenic sulfate (from DMS) Probably
important some of the time - Terrestrial biogenic sulfate (from DMS and H2S)
May play a key role in providing soluble aerosol
component - Secondary Organic Aerosol May provide a
significant part of aerosol mass, but mode of
particle formation still under question - Primary biogenic aerosol Bacteria, spores,
plant-derived particles account for a large
fraction (40-80) of aerosol mass and number !
30Smoke particles are made up mostly of
carbonaceous matter, about half of which is
water-soluble, and therefore can contribute to
CCN activity ...
31Huge variability in CCN concentrations in
Amazonia for dry and wet season
32Addition of pyrogenic CCN has pronounced impact
on cloud droplet size spectra!
33RAINDROPS Disdrometer FNS
Polluted
Clean
34Humidification and growth of biomass burning
particles
Humidification factor (shumid/sdry)
Humidification factor (ratio of wet to dry
scattering coefficients) in burning, transition
and relatively clean conditions
The preliminary results suggest that
humidification factor of biomass burning aerosols
is about 1.15. Higher values are found during the
transition period (1.3) and highest (1.4)
during the relatively cleaner condition. This
indicates increasing growth factor from dry to
wet season.
35Vertical distribution of aerosols from biomass
burning (light scattering at 500 nm and CN)
36Real time monitoring of the transport of biomass
burning emissions in South America.
CATT-BRAMS The Coupled Aerosol and Tracer
Transport model to the Brazilian developments on
the Regional Atmospheric Modeling System
The GOES-8 ABBA Fire Product on 1745Z September
7, 2002, depicting the vegetation fires on South
America. GOES resolution is 1 Km in the visible
channel, 7 and 14 Km for infrared.
The parameterized CO source emission for
September 7, 2002. Some places on Brazil with
forest biomes emitted over 2 ton km-2 of carbon
monoxide.
Source Saulo Freitas and Karla Longo
37(No Transcript)
38Another important sub-grid process,but
frequently ignored
Rondônia, 2002
Convective cloud above the Bor Forest Island
fire, 6 July 1993.
Plume-rise due to the strong buoyancy of the hot
gases/aerosols emitted
39Plume-rise of vegetation fires fires
aggregation by model host grid box
40Including plume rise mechanism troughsuper-param
eterization concept
1D plume-rise model for vegetation fires Biome
ForestTime duration 50 mnFire size 20 haHeat
flux 80 kWm-2 / 30 kWm-2
41CATT-BRAMS comparison with AIRS 500 hPa CO
Model CO (ppb) at 5.8 km without plume rise
with plume rise
22 SEP 2002
CO (ppb) from AIRS at 500 hPa
- McMillan et al., GRL 2005.
- Atmospheric InfraRed Sounder (AIRS) onboard
NASA's Aqua satellite. - CO abundances are retrieved from AIRS 4.55 ?m
spectral region.
42Aerosol Column MODIS x model
43Model CO (ppb) at 5.8 km without plume rise
with plume rise
CATT-BRAMS comparison with AIRS 500 hPa CO
CO (ppb) from AIRS at 500 hPa
25 SEP 2002
44AOT 550 nm (from biomass burning)
Solar Radiation at surface (Wm-2)
45Surface radiative forcing cooling at the surface
by -3 degree and heating at 3Km by 2.5 degrees
3 km height
3 Km
Temperature at surface
46Wet removal of PM2.5 (mg/m2) Forecast for 3 and
4/September/2004
GOESMETEOSAT IR 2100Z/4/September/2004
Wet removal of PM2.5 associated to the a
mid-latitude cold front approach
47Model comparison with near surface observations
CO and PM2.5 observations in Rondônia
48Air quality and public health
Biomass burning distribution
Exposure time () gt 80 mg/m3 2004
Exposure time () gt 80 mg/m3 2005
49São Paulo September 06, 2004
Modis image from Carlos Pires
50Campo Grande September 15, 2004
Modis image from Carlos Pires
51São Paulo September 15, 2004
Modis image from Carlos Pires
522050 BAU Scenario Deforested 2,698,735 km2
(50) Forest 3,320,409 km2 Non-forest 1,497,685
km2
33 Pg C
Soares-Filho, et al. 2006, Nature
532050 Governance Scenario Deforested 1,655,734
km2 Forest 4,363,410 km2 Non-forest 1,497,685
km2
17 Pg C
Soares-Filho, et al. 2006, Nature
54Yoram initiated all this Not everybody has
this capability Thanks for the attention!!!
55(No Transcript)
56Estimates of the Cloud Albedo radiative forcing
due to aerosols from different models
Best estimate -0.7 W/m2 Range -1.8 to -0.3
W/m2
57Estimativas da forçante radiativa direta dos
aerossóis por diferentes modelos
Best estimate -0.5 W/m2 Range -0.9 to -0.1
W/m2
58Monthly averages of column water vapor for sites
grouped by region
Schafer et al., 2007
59(No Transcript)
60Transition from pyrocu-type clouds...
61(No Transcript)
62(No Transcript)
63(No Transcript)
64(No Transcript)
65(No Transcript)
66Plume Rise Sub-grid Convective TransportCloud
venting is a very important mechanism
transporting pollutants from the PBL to the
upper levels, affecting the chemistry of
troposphere and the biogeochemical cycles.
Freitas and Longo, 2006
67Another important sub-grid process,but
frequently ignored
Rondônia, 2002
Convective cloud above the Bor Forest Island
fire, 6 July 1993.
Plume-rise due to the strong buoyancy of the hot
gases/aerosols emitted
68Model validation using CO MOPITT data for 2002
ER 100. (MOPITT MODEL)
MOPITT
350 hPa
500 hPa
700 hPa
850 hPa
69Biomass Burning or Vegetation Fires are a global
scale phenomena
Dwyer et al. 1999 J. of Biogeography
70Polluted clouds grow out of the regional haze
containing lots of CCN...
71Model comparison with CO airborne observations
from SMOCC campaign in Rondônia
Model x airborne measurements
OBSERVATION MODEL
72Plume-rise of vegetation fires typical energy
fluxes (kWm-2)
Refs Carvalho et al, 1995-2001-2005 (com.
pessoal) Riggan et al, 2004 Ward et al, 2002
Ferguson et al, 1998 Cochrane et al 200X-com.
pessoal Miranda et al, 1993.
73Cloud Condensation Nuclei at the Dry and wet
seasons in Amazonia