Effects of 2000-2050 Global Change Ozone Air Quality in the United States

1

The effects of Climate Change to the Future Air
Quality in United States
Joshua Fu, Yun-Fat Lam and Yang Gao University
of Tennessee Daniel Jacob , Loretta Mickley and
Shiliang Wu Harvard University Oct 20, 2009
2
GLOBAL CHANGE AND AIR POLLUTION (GCAP)
Joshua Fu , Yun-Fat Lam, University of
Tennessee Daniel Jacob (PI), Loretta Mickley,
Harvard University John Seinfeld, California
Institute of Technology David Streets, Argonne
National Lab David Rind, GISS/NASA
3
GCAP How will global change affect U.S. air
quality
UT
SOURCE GCAP group
4
Effect of Global Warming in GISS General
Circulation Model

• Goddard Institute for Space Studies GCM 9
  layers, 4ox5o horizontal grid, CO2 other
  greenhouse gases increased yearly from 2000 to
  2050.
• Carbon Monoxide CO
• source present-day anthropogenic
  emissions
• sink CO present-day OH fields
• Black Carbon BC
• source present-day anthropogenic
  emissions
• sink rainout

spin up
Mickley et al., 2004
SOURCE GCAP group
5
Effect of Climate Change on Regional Stagnation
GISS GCM 2 simulations for 2050 vs. present-day
climate using pollution tracers with constant
emissions
Mid-latitudes cyclones tracking across southern
Canada are the main drivers of northern U.S.
ventilation
Pollution episodes double in duration in 2050
climate due to decreasing frequency of cyclones
ventilating the eastern U.S this decrease is an
expected consequence of greenhouse warming.
Mickley et al. GRL 2004
SOURCE GCAP group
6
Climatological Fact
DECREASE IN FREQUENCY OF MID-LATITUDE CYCLONES
OVER PAST 50 YEARS
Annual number of surface cyclones and
anticylones over North America
Agee 1991
Cyclone frequency at 30o-60oN
McCabe et al. 2001
SOURCE GCAP group
7
Global model vs. Regional Model
Global Model
Regional Model
Climate
Missing aerosol feed back
Air Quality, Heat waves, Flooding, Drought, Human
Health
Direct affect solar radiation
Two-ways coupled climate and chemistry
One-way coupled climate and chemistry
8
Significant of Regional Model

• Resolution down to 1 x 1 km
• Taking advantages of detail geographical
  information in meteorology modeling, as well as
  highly reliable emission inventories for ozone
  and aerosol modeling
• All the equations in regional model are designed
  to use in fine resolution conditions
• Scalability issue in global model
• Regional/urban climate and air quality conditions
  can be simulated to provide information for local
  and regional planning
• It has better implication in model outputs

9
Development of downscaling approach

• Analysis of 2000-2050 trends in air pollution
  meteorology
• Development of GISS/GEOS-Chem interface
• Development of GISS/MM5 interface
• Development of future emission inventories for
  carbonaceous aerosols
• Application of GISS/GEOS-Chem to 2000-2050 trends
  in ozone and PM (IPCC A1B scenario)
• Statistical projection of 2000-2050 ozone trends

10
Objective

• Investigate the future air quality in United
  States for year 2050 using regional air quality
  model, CMAQ MM5
• Study the effect of global warming in regional
  scale for both climate and air quality
• Examine the effect of change of anthropogenic
  emissions
• Determine the emission reduction offsets required
  to maintain NAAQS

11
Model Configurations

• Models
• Global Models
• GISS-GCM III (GISS/NASA)
• GEOS-Chem IV (HARVARD UNIVERSITY)
• Regional Models
• MM5 and WRF (NCAR)
• CMAQ 4.6 (EPA and others)
• Interface Program Development and Regional
  Modeling (UT)
• GISS2MM5 gt MM5 (UT)
• GEOS-Chem gt CMAQ (UT)

12
Global Model Configurations
Climate Model
Chemistry Model

• GISS general circulation model III
• Global climate model
• Provide initial guess values for MM5 (Both
  current and future climate conditions - e.g. 2000
  and 2050)
• 4 x 5 horizontal resolution
• 30 vertical sigma/pressure layers

• GEOS-Chem IV
• Global chemistry model
• Provide initial and boundary conditions for CMAQ
• 2 x 2.5 horizontal resolution
• 28 vertical sigma/pressure layers
• Take into account of volcanic events, wild fire,
  lightning and dust storm across the globe

13
Regional Model Configurations
Climate Model
Chemistry Model

• MM5
• Regional climate model
• Terrain followed sigma coordination
• Resolution 108km and nest down to 36km (can be
  down to 1 km)
• 43 vertical sigma/pressure layers
• Preprocessor TERRAIN, REGRID, LITTLE_R, INTERPF
  and NESTDOWN

• CMAQ 4.6
• 14 layers (from the MM5 sigma levels)
• 36 km horizontal resolution (in this study)
• ICON and BCON from GEOS-Chem from 3 hrs to one
  hour average
• GISS/MM5 meteorological Inputs
• Input emission is compatible with 2001 EPA
  National Emission Inventory

14
GISS surface wind and temperature Inputs
Example Results of MM5 Outputs from GISS
GISS
MM5
MM5
108 km - CONUS
36 km - CONUS

• SourceL. Mickley (Harvard)

15
GISS Vs. MM5
16
GISS Vs. MM5 (JJAS)
MAX AVG Temperature
Average Temperature (K)
17
GISS Vs. MM5 (JJAS)
Temperature RMSE is below 0.4 K. And the mean
bias is close to 0.1 K ltlt 0.5
(benchmark) Wind speed RMSE is less than 0.2
m/s ltlt 2 m/s (benchmark). For the mean bias, the
value is 0.1 m/s Wind direction RMSE is less
than 20 and mean bias is less than 1 ltlt 10

18
CMAQ Simulations Scenarios

• Study period June 1 to September 1 (Ozone season)

Emission Projection
2000 NO2
2050 NO2
IPCC NOx Emission Scenario
X
19
Maximum Ozone Concentration
warmer
2000climate-2000emi
2050climate-2000emi
2000climate-2050emi
2050climate-2050emi
Fu et al. 2008
Emission has more effect than climate change on
pollution events
20
CMAQ Simulations Output (JJAS)
MAXIMUM OZONE
AVERAGE OZONE
21
Source Harvard University
22
GEOS-Chem Vs. CMAQ (JJAS)
23
GEOS-Chem Vs. CMAQ (JJAS)
24
GEOS-Chem Vs. CMAQ (JJAS) MidN
Large difference in temperature
Small difference in wind speed
The CMAQs trend is similar as GEOS-Chems trend
(Temperature dominated case)
25
GEOS-Chem Vs. CMAQ (JJAS) - NE
Small difference in temperature
Small difference in wind speed
The CMAQs trend is not similar as GEOS-Chems
trend (Emission domination case)
26
GEOS-Chem Vs. CMAQ (JJAS) SE
Large difference in temperature
Large difference in wind speed
The CMAQs trend is not similar as GEOS-Chems
trend (Cloud dominated case)
?? Why different from NE ??
27
Remarks

• Global downscaling of GISS and GEOS-Chem have
  successfully performed with high confidence.
• The ozone trend (2050 2000) of GEOS-Chem and
  CMAQ are found to be quite difference, where
  GEOS-Chem is much more temperature driven
• (may due to the coarse resolution of
  meteorological data)
• In GEOS-Chem, climate change is a stronger factor
  than emission change for MidN and NE, but not
  showing in SE
• In CMAQ, only MidN have shown stronge climate
  change effect.
• Overall, the maximum zonal ozone concentration in
  2050 is much higher than 2000. However, the
  probability of getting higher ozone may not
  higher. The convection and cloud cover have
  played important role on this issue.

28
THANK YOU!
29
Average Aerosol (PM2.5) Concentration
Emission has more effect than climate change on
pollution events ?
BLACK CARBON
SULFATE AEROSOL
Climate Effect
Emission Effect
US
US
Climate change doesnt effect South-East ?