Transcript PPT - Atmospheric Chemistry Modeling Group

Effect of climate change on air pollution episodes in the
United States: a model study
Loretta J. Mickley, Daniel J. Jacob, Brendan D. Field
Harvard University
David Rind
Goddard Institute for Space Studies
We know that day-to-day meteorology affects the severity and duration of
pollution episodes.
New England
days
Number of summer days with 8-hour ozone
> 84 ppbv, average for northeast U.S. sites
1988, hottest
on record
Probability
of ozone exceedance
vs. daily max. temperature
Lin et al. 2001
Why does probability of ozone episode increase with increasing temperature?
Faster chemical reactions, increased biogenic emissions, and stagnation.
How will a changing climate affect pollution?
Answer: we don’t know.
Rising temperatures could mean faster chemical reactions. . .
Higher surface temperatures could also mean a deeper boundary layer,
diluting concentrations at the surface.
The picture is complicated.
Top of boundary layer
Soup of
pollution
precursors
{
ozone, aerosol
strong mixing
How to make pollution:
Need sunlight, water vapor, and a mix of
anthropogenic or natural “ingredients.”
H2O
Hydroxyl (OH)
winds
Ozone (O3)
+
Nitrogen oxides
CO, Hydrocarbons
rainout
(important for
aerosols)
deposition
Fires
Biosphere
Human
activity
Our approach: focus on changes
in winds and rainout.
Previous studies have focused mainly on chemical response to
temperature change (e.g. Aw and Kleeman, 2003)
Increase in surface ozone
and aerosol due to 5K
temperature change
DO3
all other met variables --e.g.
circulation, boundary layer
height-- the same
Ozone increase 10-15%
due to faster reaction rates.
DAerosol
Aerosol decreases 10-15%
due to increased
volatilization of ammonia.
What have long-lived tracer studies shown about changes in transport?
DSF6 surface
Rind et al., 2001
31-layer GISS GCM, several longlived tracers, 2xCO2
Increased convection leads to:
less surface SF6
and
DSF6 500 mb
more SF6 aloft.
Holzer and Boer, 2001
coupled global model, 2xCO2
Less vigorous flow, increased
plume concentrations
How to make pollution:
Need sunlight, water vapor, and a mix of
anthropogenic or natural “ingredients.”
H2O
Hydroxyl (OH)
winds
Ozone (O3)
+
Nitrogen oxides
CO, Hydrocarbons
rainout
(important for
aerosols)
deposition
Fires
Biosphere
Human
activity
Our approach: focus on changes
in winds and rainout.
Pilot Project: Implement “tracers of anthropogenic pollution” into GISS
General Circulation Model
Timeline
1950
spin-up (ocean adjusts)
2000
increasing A1 greenhouse gas
2050
Goddard Institute for Space Studies GCM: 9 layers, 4ox5o horizontal grid, CO2
+ other greenhouse gases increased yearly from 2000 to 2050.
July global mean temperature
Carbon Monoxide: COt
source: present-day anthro emissions
sink: CO + present-day OH fields
2045-2052
+2o C Temp change
spin up
Sensitive to climate change
Circulation also sensitive to climate change
{
Black Carbon: BCt
source: present-day anthro emissions
sink: rainout
19952002
Anthropogenic emissions:
• What changes:
Well-mixed greenhouse gas
concentrations over time
Climate response to greenhouse
gas trends, including rainout of
black carbon tracer
CO emissions (molecules /s)
• What remains the same:
Emissions of CO and black
carbon tracers
Sink of CO (monthly mean,
present-day OH)
BC emissions (kg/s)
Timeline
1950
spin-up (ocean adjusts)
2000
increasing A1 greenhouse gas
2050
Large-scale meteorological changes from 2000 to 2050, Jul-Aug mean
D Surface temp
D Surface pressure
D Precipitation
Increased surface temperature
co-located with decreased surface
pressure in mid-U.S.
Decreased precipitation in
Southeast. (20-30% less in some
gridboxes)
How much do July-August mean COt mixing ratios change at the surface?
present-day 1995-2002
future 2045-2052
Answer: not much
Not much change either for BCt.
Note: only source of COt is direct
emission
Difference 2050-2000
Our approach: Look at daily mean concentrations averaged over
specific regions for two 8-year intervals (1995-2002) and (2045-2052).
Histogram of COt concentrations
averaged over Northeast for
1995-2002 summers (July-Aug)
midwest
California
northeast
southeast
Cumulative probability plot shows the
percentage of points below a certain
concentration.
Frequency distributions for surface COt and BCt show significantly
higher extremes in 2050s compared to present-day.
July - August
2045-2052
1995-2002
Changes at the extremes are due solely to changes in circulation and rainfall.
Frequency distributions for three U.S. regions in July-August show increased
severity of pollution episodes.
2050
2000
In all regions, daily COt and
BCt concentrations correlate
(R2 ~ 0.6 – 0.8) so much of the
difference is likely due to
circulation.
How does depth of boundary layer change with changing climate?
Northeast daily maximum
boundary layer height.
Triangles indicate days of high pollution.
2045-2052
Extreme pollution events associated
with lower boundary layer heights.
1995-2002
Higher BL heights in future go in
opposite direction to what is needed to
explain air quality differences.
Evolution of a typical pollution event. This happens repeatedly
during summertime.
weak winds
cyclone (low pressure system)
BCt and wind
fields for 6
consecutive days
in summer.
cold front from
Canada
100 x mg/m3
Is pollution more persistent in future? How often do cold fronts come
through to sweep away pollution?
Mean frequency of cold fronts
pushing into Midwest decreases
by ~20% in future climate.
Persistence of pollution
episodes increases by
30-100% over Midwest.
Cyclone number and cold
front frequency decline in
future, allowing pollutants
to build up.
A decrease in cyclone frequency over midlatitudes has also been
observed in recent decades.
1000
cyclones
Agee, 1991
500
100
1950
anticyclones
1980
annual number of surface
cyclones and anticylones for
North America and nearby
ocean
McCabe et al., 2001
30-60N
Standardized departure of
cyclone frequency over
Northern Hemisphere.
Model studies of future climate have
found similar declines relative to the
present-day. (e.g., Zhang and Wang, 1997;
Carnell and Senior, 2001; Geng and Sugi,
2003)
Two mechanisms for the meridional transport of energy on a round,
wet world.
1. Mid-latitude cyclones push warm
air poleward ahead of front, push
cold air equatorward behind front.
warm tropics
cold poles
cold front
2. Eddy transport of latent heat
carries energy to higher latitudes.
Reasons for decline in cyclone generation over midlatitudes.
DT
Change in zonally averaged
temperature for July-August.
Increase is greatest at high
latitudes. Reason is ice-albedo
feedback.
Change in northward transport of
latent heat by eddies in midtroposphere in future atmosphere.
Reduced temperature gradient
and more efficient eddy transport
of energy poleward
Fewer cyclones generated
More persistent pollution
events
How do you translate our results into “ozone alert days”?
Model predicts high-pollution days will occur about 66% more frequently in future due to
changes in circulation over Northeast and Midwest.
Best calculation includes full chemistry responding to all the meteorological changes.
Hotter maximum temperatures
Triangles
indicate days of
highest BCt
concentrations.
2050
2000
Reduced cloud cover
High maximum temperatures and reduced cloud cover suggest increased
ozone production, amplifying effect of stagnation.
GCAP: ongoing project to diagnose effect of changing climate on
US air quality
Global Climate and Air Pollution
collaborators: David Streets, John Seinfeld, David Rind, Joshua Fu
23L GISS GCM, with changing GHGs
1950
2000
Spin-up of ocean
2025
2050
2075
2100
archived
temperatures,
humidity, winds, etc
GEOS-CHEM
Calculate chemistry, aerosol
present-day & future precursor
emissions
CMAQ
regional
model
Monitoring pollution and biomass burning over North America with satellites
AIRS instrument onboard the AQUA satellite enables observation
of complex and overlapping long-range transport.
AIRS CO Column July 18, 2004
fires
Asian pollution
GEOS-CHEM CO Column July 18, 2004
model
U.S. pollution
Wallace McMillan (UMBC)
Solene Turquety (Harvard)
First day of ozone column data from TES
TES = Tropospheric Emission
Spectrometer
Measures infrared radiances in
both limb and nadir mode.
Launched July 15, 2004
pollution
Will provide a detailed, global
view of ozone, CO, and HNO3
pollution
First day of data!
biomass burning
Tropospheric ozone column on
September 20, 2004
Summary
Model predicts an increase in the severity and duration of pollution
episodes over the Midwest and Northeast U.S. by 2050, even with
constant emissions.
Change in pollution tied to a decrease in the frequency of cold fronts
arriving from Canada, which sweep away the pollution.
2050s
Observed correlations between
meteorological parameters and pollutant
concentrations provide a tool for
predicting trends in GCM simulations.
A new era of satellite observations probing
the troposphere can supply data to
assess our model predictions.
2000s
Extra slides
Meteorology in NE associated with high CO and BC events includes
average to high temp and pressures, low winds, and clear skies
Low trop clouds
2050
2000
Surf temp
Surf press
Surface wind
speed
Boundary
layer height
Static stability
Days of
high CO
Present-day 1995-2002
Future 2045-2052
July-August mean BC
concentrations at the surface
Change from present-day