Interactions between climate and atmospheric chemistry in the US

1
Interactions between climate and atmospheric
chemistry in the US

• Effects of climate change on air quality
• Effects of short-lived species on climate

Smog over Pittsburgh, ranked 1 city for
particulate pollution in 2008 by American Lung
Association
Loretta J. Mickley, Harvard University Collaborat
ors Rynda Hudman, Daniel Jacob, Eric
Leibensperger, Jennifer Logan, Havala Pye,
Dominick Spracklen, Amos Tai, Shiliang Wu,
Moeko Yoshitomi Funding for this work NASA, EPA,
EPRI
2
Part 1 Effects of climate change on air quality.
3
Millions of people in U.S. already live in areas
of high pollution. How will a changing climate
affect pollution?
Calculated with new 0.075 ppm standard
Number of people living in areas that exceed the
national ambient air quality standards (NAAQS) in
2008.
4
O2
hn
Chemistry of tropospheric ozone oxidation of CO,
VOCs, and methane in the presence of NOx
O3
STRATOSPHERE
8-18 km
TROPOSPHERE
Stagnation promotes ozone production
hn
NO2
NO
O3
hn, H2O
OH
HO2
H2O2
Deposition
CO, VOC
Nitrogen oxide radicals NOx NO NO2
combustion, soil, lightning Methane wetlands,
livestock, natural gas Nonmethane volatile
organic compounds (VOCs) vegetation,
combustion, industry CO (carbon monoxide)
combustion, VOC oxidation
Tropospheric ozone precursors
5
Weather plays a large role in ozone air quality.
The total derivative dO3/dT is the sum of
partial derivatives (dO3/dxi)(dxi/dT). x
ensemble of ozone forcing variables that are
temperature-related.
Lin et al., 2001
6
Low pressure systems (aka cyclones) cross
southern Canada and sweep out ozone pollution
from Eastern US.
cold front
EPA ozone levels
L

• Stalled high pressure system associated with
• increased biogenic emissions
• clear skies
• weak winds
• high temperatures.

Hazardous levels of ozone
3 days later
cold front
L
Cold front pushes smog poleward and aloft on a
warm conveyor belt.
7
Cyclone passage through southern Canada/Great
Lakes region strongly affects frequency and
duration of U.S. ozone episodes.
Correlation between cyclone number each summer in
red and green boxes and number of US ozone
episodes
Sample storm tracks, summer 1979-1981
27 year record
Strong anti-correlation of cyclone number and
number of ozone episodes in eastern US
Fewer cyclones per summer in green box
leads to more ozone episodes in US.
Leibensperger et al., 2008
8
1950-2000 observed trend in cyclone frequency
matches that in climate model with increasing
greenhouse gases.
1950-2006 trend in JJA cyclones in S. Canada
Trend in cyclones appears due in part to weakened
meridional temperature gradients, reduction of
baroclinicity over midlatitudes. What does this
trend mean for ozone pollution in US? Emissions
of ozone precursors have declined during this
period.
NCEP/NCAR obs
Mickley et al., 2004 Leibensperger et al., 2008
9
Trend in emissions and trend in cyclones have
competing effects on surface ozone.
Cyclones less frequent cyclones cold fronts
mean more persistent pollution episodes Emissions
reduced emissions means fewer episodes.
Decline in emissions of ozone precursors from US
mobile sources. Parrish 2006.
Mickley et al., 2004 Leibensperger et al., 2008
10
Ozone pollution days in the Northeast US
We find that if 1980-2006 cyclone frequency had
remained constant, we would have had zero
episodes over Northeast.
If emissions had remained constant, decline in
mid-latitude cyclone number over Canada would
have meant more persistent stagnation episodes,
more ozone.
Climate response
Trend in pollution days due to decline in cyclone
frequency
days yr-1
days yr-1
Trend in pollution days due to decline in
emissions
days yr-1
11
Particulate matter (PM, aerosols) sources and
processes
ultra-fine (lt0.01 mm)
fine (0.01-1 mm)
cloud (1-100 mm)
precursor gases
oxidation
nucleation
cycling
coagulation
H2SO4
SO2
condensation
RCO
VOCs
coarse (1-10 mm)
scavenging
NOx
HNO3
wildfires
NH3
carbonaceous combustion particles
combustion biosphere volcanoes
soil dust sea salt
agriculture biosphere
12

• Observed correlations of total PM2.5 with
  meteorology
• Precipitation
• Stagnation
• Temperature
• Positive correlation with temperature occurs due
  to
• Increased oxidation of SO2
• Greater biogenic emissions

Precipitation
Stagnation
Results from EPA AQS database 1000 sites
sampled every 1-6 days from 1998 to 2007.
Observed correlations provide means to test
model simulations.
Temperature
Tai et al., ms.
13
What do models project for future air quality? We
have developed GCAP (Global Change and Air
Pollution).
GISS GCM Physics of the atmosphere Qflux ocean,
well-mixed GHGs
GEOS-Chem Chemical transport model chemistry,
emissions
met fields
met fields
chemistry fields
met fields
Regional chemistry model
Regional climate model
Chemistry model driven by GCM meteorology to
study influence of climate on air quality.
Mickley et al., 2004
14
2000-2050 change in max daily 8-hour average JJA
ozone
How will US surface ozone change in a changing
climate?
Climate penalty for air quality Harvard model
shows 2-12 ppb increase in surface ozone in East
Most models agree that surface ozone will
increase over the Northeast. Disagreement occurs
elsewhere due to differences in chemistry and
cloud cover change.
ppb
Multi-model comparison Weaver et al., 2009
15
Uncertainty in response of surface PM to changing
meteorology is large. We can use present-day
observations to test models.
Calculated response in surface PM to 2.5 oC
temperature change applied uniformly for
July. Dawson et al., 2007
(µg/m3)
Observed correlation between surface temperature
and surface PM concentrations.

• Positive correlation with T due to
• Increased oxidation of SO2
• Greater biogenic emissions

Tai et al, ms. in progress
16
Part 2 Effects of short-lived species on
climate.Case study of US aerosols and regional
climate change.

• Radiative forcing
• Easily calculated metric of climate change
• Suggests the relative magnitude of surface
  temperature response to a given perturbation.

17
Present-day radiative forcing due to aerosols
over the eastern US is comparable in magnitude,
but opposite in sign, to global forcing due to
CO2.
Due to short lifetime, forcing due to aerosols is
not uniform across globe. Over the US, radiative
forcing due to sulfate aerosols is -2 Wm-2.
cooling
Globally averaged radiative forcing due to CO2 is
1.7 Wm-2. warming
IPCC, 2007 Liao et al. , 2004
18
Trend in aerosols over United States suggests
cleaner skies, possible warming?
Calculated trend in surface sulfate
concentrations, 1950- 2001.
Sequence shows increasing sulfate from 1950-1980,
followed by a decline in recent years.
Comparison to observed sulfate concentrations
shows good agreement.
Leibensperger et al., ms
19
Is the climate response to changing aerosols
regional or global?
Recent US Climate Change report suggests more
global than regional response, but the report
looks at an ensemble of short-lived species all
over the globe. U.S. Climate Change Science
Program, Synthesis and Assessment Product 3.2
Harvards work to date suggests more regional
than global response at least for US
aerosols. Decline in the aerosol burden over the
eastern US will lead to regional warming, in a
way that the US Climate Change report would not
have recognized.
Calculated present-day aerosol optical depths
20
What is the influence of changing aerosol on
regional climate? In pilot study, we zero out
aerosol optical depths over US.
GISS GCM
For pilot study, 2 scenarios were simulated
Control aerosol optical depths fixed at 1990s
levels. Sensitivity U.S. aerosol optical
depths set to zero (providing a radiative forcing
of about 2 W m-2 locally over the US)
elsewhere, same as in control simulation. Each
scenario includes an ensemble of 3
simulations. Caveats No transport, only direct
effect considered in this pilot study.
21
Removal of anthropogenic aerosols over US
increases annual mean surface temperatures by
0.5 o C. Summertime temperatures increase as
much as 1.5 oC.
Warming due to 2010-2050 trend in greenhouse
gases.
Additional warming/ cooling due to zeroing of US
aerosols
oC
oC
Mean 2010-2050 temperature difference
No-US-aerosol case Control White areas signify
no significant difference. Results from an
ensemble of 3 for each case.
Annual mean surface temperature change in
Control.
Mickley et al., ms. 2009
22
The regional surface temperature response to
aerosol removal persists for many decades in the
model.
Annual mean temperature trends over Eastern US
Bottom line Efforts to clear the air of
anthropogenic aerosol over the US may exacerbate
regional warming.
Mickley et al., ms
23
Ongoing study Perform realistic simulation of
changing aerosol optical depths over the US,
together with sensitivity studies.

• We use historical/projected emissions of SO2,
  NOx, BC, and OC to quantify the climatic role of
  US aerosols in the past and future.
• 1950-2050 Control simulation (EDGAR/Tami Bond
  historical emissions and A1B includes rising
  U.S. aerosol sources until 1980 and subsequent
  decline)
• Sensitivity simulations
• 1950-2050 No US aerosols.
  
• Quantifies the past effect of
  U.S. anthropogenic sources on regional climate.
• 2010-2050 Constant US emissions
  Quantifies the warming effect from the projected
  decrease in U.S. emissions.

GEOS-Chem chemistry transport model
aerosol concentrations
Calculation of cloud droplet number concentrations
aerosol indirect effect
GISS GCM III climate model
Climate response to aerosol trends over the US
24

• Implications for policymakers
• Policymakers need to consider climate change
  penalty, i.e., the additional emission controls
  necessary to meet a given air quality target.
• Efforts to clear the air of anthropogenic
  aerosol over the US may exacerbate regional
  warming.

Directions for future research Understand causes
in interannual variability of air
quality. Investigate model sensitivity of
pollutants to meteorology, and compare to
observations. Understand chemistry of biogenic
species, e.g. isoprene Improve emission
inventories for recent past/future, especially
for NH3, black carbon, organic carbon,
mercury Understand secondary organic aerosols
sources, chemistry. Improve modeling of fine
scale features, investigate how best to downscale
meteorology from global climate models, test
effects of land use change. Understand
aerosol-cloud interactions, characterize aerosol
composition
25
Extra slides
26
Observed Correlations of PM2.5 with Meteorology
Multiple linear regression to fit 1998-2008
deseasonalized EPA/AQS data for PM2.5

• Observed ve correlation with sulfate is larger
  by 10x ? stagnation and air mass origins
• Strong ve correlation with nitrate in the west
  and north ? agricultural NH3 and soil NOx
  emissions
• Strong ve correlation with OC ? VOC emissions
  and fires

27
2000-2050 climate change increases JJA surface
ozone 1-5 ppb on average across US, 5-10 ppb
during heat waves in Midwest
Max. 8-hr-avg ozone
Effect of climate change alone
2000s conditions 2050s climate 2050s
emissions 2050s climate emis
Increase of summer max-8h-avg ozone
99th percentile
Cumulative probability ()
Midwest
Daily max 8h-avg ozone averaged in JJA (ppb)
We define the climate change penalty as the
effort required to meet air quality standards
under future climate change.
Wu et al., 2007
28
We define the climate change penalty as the
effort required to meet air quality goals in the
future atmosphere.
40 cut in NOx 2050s climate
present-day NOx emissions climate

climate change penalty
40 cut in NOx present-day climate
50 cut in NOx 2050s climate
20002050 climate change implies an additional
25 effort in NOx emission controls to achieve
the same ozone air quality.
Midwest surface ozone
Wu et al., 2007
29
2000-2050 JJA surface temperature change
Models tend to agree on 2000-2050 changes in
surface temperature over the United States.
Most models agree that surface ozone will
increase over the Northeast. Disagreement occurs
elsewhere due to differences in chemistry and
cloud cover change.
Weaver et al., 2009
30
Change in annual mean surface inorganic aerosol
from 2000-2050 climate change (no change in
emissions)
Increase in Northeast due to increased
temperature and accelerated oxidation rates
Decrease in Southeast due mainly to increased
precipitation.
Calculation of future aerosol levels is
challenging because of uncertainty in future
rainfall over mid-latitudes. Also, mix of aerosol
species is expected to change, so sensitivity to
climate will also change.
Present-day annual average
Pye et al., 2009
sulfate nitrate ammonium
31
Projected increase in wildfires could affect air
quality in the US.
We predict future wildfires using observed
relationships between meteorology and area burned
for different ecosystems.
2000-2050 changes in fire season surface ozone.
Spracklen et al., 2009 Hudman et al, ms.
32
Projected increase in wildfires could affect air
quality in the US.
2000-2050 change in JJA surface organic aerosol
due to increased wildfires
We have developed a fire prediction tool based on
observed relationships between meteorology and
area burned. Applying these relationships to GCM
meteorology, we predict area burned and future
emissions of wildfire pollutants.
mg m-3
Changes in JJA surface ozone concentrations
Spracklen et al., 2009 Hudman et al, ms.
33
Observations of a possible relationship between
trends in aerosol optical depths and surface
temperature.
Smoothed monthly mean AOD for sites in Europe
Annual mean fluxes, temperatures
Ruckstuhl et al., 2008