SEARCH

SEARCH BY CITATION

Keywords:

  • global climate change;
  • ozone;
  • fine particulate matter

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results and Discussion
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

[1] Simulated future summers (i.e., 2049–2051) and annual (i.e., 2050) average regional O3 and PM2.5 concentrations over the United States are compared with historic (i.e., 2000–2002 summers and all of 2001) levels to investigate the potential impacts of global climate change and emissions on regional air quality. Meteorological inputs to the CMAQ chemical transport model are developed by downscaling the GISS Global Climate Model simulations using an MM5-based regional climate model. Future-year emissions for North America are developed by growing the U.S. EPA CAIR inventory, Mexican and Canadian emissions and by using the IMAGE model with the IPCC A1B emissions scenario that is also used in projecting future climate. Reductions of more than 50% in NOX and SO2 emissions are forecast. Impacts of global climate change alone on regional air quality are small compared to impacts from emission control-related reductions, although increases in pollutant concentrations due to stagnation and other factors are found. The combined effect of climate change and emission reductions lead to a 20% decrease (regionally varying from −11% to −28%) in the mean summer maximum daily 8-hour ozone levels (M8hO3) over the United States. Mean annual PM2.5 concentrations are estimated to be 23% lower (varies from −9% to −32%). Major reductions in sulfate, nitrate and ammonium PM2.5 components combined with the limited reduction in organic carbon suggests that organic carbon will be the dominant component of PM2.5 mass in the future. Regionally, the eastern United States benefits more than the rest of the regions from reductions in both M8hO3 and PM2.5, because of both spatial variations in the meteorological and emissions changes. Reduction in the higher M8hO3 concentrations is also estimated for all subregions and fewer days with M8hO3 above the air quality standards in urban sites with Atlanta in the southeast benefiting most.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results and Discussion
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

[2] Recent observations and future projections suggest that regional air quality will respond to global climate change and the two systems are intrinsically coupled [e.g., Intergovernmental Panel on Climate Change (IPCC), 2001; National Research Council (NRC), 2001; Brasseur and Roeckner, 2005]. However, our understanding of the linkages between air quality and climate change remain incomplete, in part, because of the disparate spatial and temporal scales traditionally used in the study of these fields.

[3] Climate change over the next century is predicted to have a direct impact on meteorology [IPCC, 1996]. Leung and Gustafson [2005] discuss the potential for air quality changes in the western and southwestern United States in the 2050s based on changes in surface air temperature, downward solar radiation, precipitation frequency, stagnation events and ventilation in future climate simulations for the United States. Mickley et al. [2004] suggest the reduced cyclone frequency in a future warmer climate could increase the severity of summertime pollution in the northeastern and Midwestern United States, although the increase of hurricane strength and precipitation might counteract this in some regions [Webster et al., 2005]. Hogrefe et al. [2004] estimate that regional climate change alone will increase the summertime average daily maximum 8-hour ozone concentration over the eastern United States by 4 ppb in the 2050s. Their results are based on the IPCC A2 emission scenario [IPCC, 2000], which is one of the highest future emissions scenarios. Knowlton et al. [2004] estimate that in 2050 there will be a 4.5% increase in O3-related acute mortality in New York metropolitan area, although some researchers [Schwartz et al., 2005] question their findings. Using a similar approach, Murazaki and Hess [2006] estimate 0–2 ppb decreases in U.S. background ozone and an increase up to 6 ppb within the United States in 2100 compared to 2000 due to climate change alone. Recently, Langner et al. [2005] have examined the impact of global/regional climate change on surface ozone and deposition of sulfur and nitrogen in Europe. A strong increase in surface ozone and mean of daily maximum over southern and central Europe and a decrease in northern Europe have been estimated. The decrease in wet deposition of sulfate and nitrate over western and central Europe is caused by the reduction in precipitation, but the authors caution that longer simulation periods are necessary to establish the changes in deposition.

[4] Recent studies suggest that, at least for ozone, future pollutant concentrations are more sensitive to the expected changes in precursor emissions than to the expected changes in temperature and photolytic flux [e.g., Bergin et al., 1999; Russell and Dennis, 2000]. As the amount of ozone formed per NOX molecule emitted remains somewhat constant [e.g., Kleinman, 2000], except in areas with very high emissions of NOX [e.g., Ryerson et al., 2001] or high in reactive VOC emissions and given the small change estimated for VOC emissions, forecast differences in ozone will generally depend on the forecast NOx emission changes. This should not be construed as suggesting that ozone will not respond to VOC controls as significant evidence suggests otherwise, even in cities with high biogenic loadings [Cohan et al., 2006]. Global/regional climate change will have relatively less impact on NOX emissions since they are largely anthropogenic and they do not show a strong function of temperature. Thus current policies in the United States to reduce NOX emissions, such as those being pursued now (http://www.epa.gov/ttn/chief/trends/index.html) should continue to be effective.

[5] The objective of this study is to assess the impacts of global climate change on regional air quality over the United States. Here, both the direct (impact of climate change on meteorology) and indirect impacts (those caused by emission changes due to either/both controls and climate change) are evaluated. We focus on O3 and fine particulate matter (FPM) because of their suspected significant human health effects [e.g., Pekkanen et al., 1997; Galizia and Kinney, 1999; El-Fadel and Massoud, 2000; NRC, 2001]. Specifically, we follow PM2.5 (PM with an aerodynamic diameter less than 2.5 μm). Future O3 and PM2.5 concentrations are compared to historic ones under two different cases: In the first case, impacts of changes on regional air quality in the United States by climate change alone are examined by keeping emissions sources, activity levels and controls constant. In the second case we estimate the future pollutant concentrations based on changes in climate and emissions using the IPCC A1B emission scenarios [IPCC, 2000] and planned controls.

2. Methods

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results and Discussion
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

[6] Air quality modeling was conducted using the Community Multiscale Air Quality (CMAQ) Modeling System [Binkowski and Roselle, 2003] and meteorology downscaled from the Goddard Institute of Space Studies (GISS) Global Climate Model (GCM) [Rind et al., 1999] using the Penn State/NCAR Mesoscale Model (MM5) [Grell et al., 1994]. Future-year emissions forecast for North America are developed by forecasting activity growth and application of emission controls, as discussed below.

2.1. Emissions

[7] The 2001 Clean Air Interstate Rule (CAIR) emission inventory (EI) (http://www.epa.gov/cair/technical.html) is used as the U.S. emission inventory for the historic period (i.e., 2000–2002), as well as the basis for projected emissions up to 2020. For Canada, the Environment Canada (EC)'s 2000 inventory has been used for area and mobile sources (http://www.epa.gov/ttn/chief/net/canada.html). For point sources, the 2002 inventory that the New York State Department of Environmental Conservation compiled using National Pollution Release Inventory (NPRI) was scaled using EC's state level summary. For Mexico, the U.S. EPA's 1999 BRAVO inventory has been updated with the Mexico NEI (http://www.epa.gov/ttn/chief/net/mexico.html).

[8] Projection of emissions is done in two steps: (1) For near future (2001–2020), the 2020 CAIR EI of the U.S. EPA is modified using Economic Growth Analysis System (EGAS) factors (http://www.epa.gov/ttn/ecas/egas5.htm) and (2) for far future (2020–2050) projections are carried out on the basis of the Netherlands Environmental Assessment Agency's IMAGE model (http://www.mnp.nl/image), which uses widely accepted scenarios (i.e., Intergovernmental Panel on Climate Change (IPCC) Special Report on Emissions Scenarios (SRES)) [IPCC, 2000]. The scenario SRES-A1B has been selected for the far future projection in order to be consistent with the climate/meteorological modeling used here. The SRES A1B scenario describes a future world of rapid economic growth and global population. Emissions peak in midcentury and decline thereafter because of rapid introduction of more efficient technologies, and balanced usage between fossil fuels and other energy sources. Emissions are processed by the Sparse Matrix Operator Kernel Emissions (SMOKE v2.1) Modeling System (http://cf.unc.edu/cep/empd/products/smoke/index.cfm). SMOKE converts the resolution of the data in an emission inventory to the resolution needed by the air quality model. Emission inventories are divided into the following source categories: area sources, nonroad mobile sources, on-road mobile sources, point sources and biogenic land use data. MOBILE6 is selected for mobile source emissions (http://www.epa.gov/OMS/m6.htm). The BELD3 land use database (http://www.epa.gov/ttn/chief/emch/biogenic) is used for estimating biogenic emissions, and is not modified between the historic (i.e., 2001) and future (i.e., 2050) cases because of the lack of information. Historic and future emission inventories include the following compounds: carbon monoxide (CO), nitrogen oxides (NOX), sulfur dioxide (SO2), nonmethane volatile organic compounds (NMVOC), ammonia (NH3), and speciated particulate matter (PM10 and PM2.5). A detailed description of the method has been presented by Woo et al. [2006].

2.2. Meteorology

[9] Meteorological fields are derived from the GISS GCM [Rind et al., 1999], which was applied at a horizontal resolution of 4° latitude by 5° longitude to simulate current and future climate at global scale [Mickley et al., 2004]. The simulation followed the SRES-A1B emission scenario [IPCC, 2000] for greenhouse gases. Note that for consistency, the same emission scenario is used in projecting future emissions described in 2.1. Leung and Gustafson [2005] downscaled the GISS simulations for 1995–2005 and 2045–2055 using the Penn State/NCAR Mesoscale Model (MM5) [Grell et al., 1994] to the regional scale. MM5 is applied in a nested configuration with 108 km horizontal resolution for the outer domain and 36 km for the inner one. The inner domain covers the continental United States, part of Canada, Mexico and ocean (Figure 1). The Meteorology Chemistry Interface Processor (MCIP) (http://www.cmascenter.org) is used to provide the meteorological data from the hourly MM5 outputs needed for the emissions and air quality models that both have 147 × 111 horizontal grids of 36 km × 36 km, with nine (9) vertical layers up to approximately 15 km.

image

Figure 1. Modeling domain and regions examined.

Download figure to PowerPoint

2.3. Air Quality Modeling

[10] Using meteorology simulated by MM5, both a full historic (2001) and future year (2050) as well as three summer (June-July-August) episodes for historic (2000–2002) and future (2049–2051) O3 and PM2.5 concentrations are simulated using the CMAQ Modeling System with the SAPRC 99 chemical mechanism. Predicted pollutant (i.e., O3 and PM2.5) concentrations for the historic periods are compared with the observed in order to evaluate the modeling system performance. For the future period two different cases are examined. In the first case the same emission state, i.e., the 2001 inventory, is used for both historic and future simulations in order to estimate the impact to air quality by changes in global climate alone. Although the emission inventory is kept the same, emissions are not, since some pollutant emissions (e.g., biogenic and mobile sources) depend on meteorology. In the second case the combined impact of future emissions (based on the forecast emissions and climate) and future climate is evaluated to simulate future levels of O3 and PM2.5. Average regional concentrations are predicted for the United States and five subregions (Figure 1). In this work, changes in long-range transport of pollutants to the United States have been neglected as these are uncertain and could mask the impacts of processes investigated here. In both historic and future periods, boundary conditions are kept the same, as there is insufficient information for the emission scenario we use. Keeping the boundary conditions constant makes the impact of regional climate change on pollutant concentrations more transparent. Given the simulated small sensitivity of air quality to climate change, imposing varying boundary conditions would add significant noise to our ability to isolate how climate change impacts compared to emissions changes.

3. Results and Discussion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results and Discussion
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

3.1. Emissions

[11] Emissions changes between future (2050) and historic (2001) years show large decreases in SO2 (−51%) and NOX (−51%) when climate change, growth in human activities and emission controls are simulated (2050 emission inventory and 2050 meteorology) (Figure 2). These reductions are due to control strategies applied to anthropogenic U.S. and Canadian sources while the growth of the industrial sector gives higher emissions in Mexico. Emission reductions in anthropogenic VOCs combined with the higher biogenic emissions in the warmer climate results in a small change in VOC emissions (+2%). A detailed description of the regional emissions has been presented by Woo et al. [2006]. For the case where only climatic changes are considered, VOC emissions are higher (+15%) in the future because of temperature effect on biogenic and mobile sources. Minor increases in NOX (+2%) and SO2 (+4%) are also predicted.

image

Figure 2. Yearly emissions for 2001, 2050 and 2050 for the “no emissions projection” scenario (2050_np).

Download figure to PowerPoint

3.2. Meteorology

[12] Meteorological model performance is evaluated by comparing hourly statistical distributions of observed and predicted temperatures over the U.S. (Figure 3) data from more than 1000 monitoring stations (see http://dss.ucar.edu/datasets/ds472.0/). Leung and Gustafson [2005] provide details, though a summary is given here. There is a small warm bias of 0.4 K in the average summer temperatures of 2000–2002. Model performance is better for the northeast region with a small cold bias of 0.1 K, and poorest for the southeast, with a warm bias of 1.5 K. A general cold bias in the 2001 annual temperature is found. Cumulative distribution function (CDF) plots presented in auxiliary material (Figures 1, 2, 3 aux. mater.) compare observed and predicted temperatures for 3 consecutive years (2000, 2001, 2002). A general underprediction is observed in all subregions but the model tends to overpredict maximum temperatures. Model performance is better during the summer months and worst during the transition season of fall, when mesoscale variability is high. As discussed by Leung and Gustafson [2005] data assimilation has not been used. Previously MM5 evaluations [e.g., Zhang et al., 2005] reveal that it reproduces well the diurnal variations for temperature and relative humidity (RH), and the minimum temperatures. It tends to overpredict maximum temperatures and underpredict both maximum and minimum RHs. Moreover, MM5 predicts well the wind speeds but poorly the wind direction and the maximum mixing depths.

image

Figure 3. Mean summer (2000–2002) and mean annual (2001) observed and predicted temperatures and monthly standard deviations.

Download figure to PowerPoint

[13] Future summer temperatures (i.e., 2049–2050) compared to the historic ones (i.e., 2000–2002) are simulated to be 1.4 K warmer in the United States (Figure 4), with small variations by region (±0.6 K). The minimum increase is noted in the Midwest (0.8 K) and the maximum in the west (2.0 K). The 2050 annual average temperature is simulated to be 1.7 K warmer than 2001 in the United States, with small variations by region (±0.5 K). Maximum warming occurs during fall with simulated average temperature changes up to 4.8 K in the west. The standard deviation calculated on the monthly average temperatures is higher for the annual simulation compared to summers in both observations and predictions. This is caused by the higher variation in temperature during a whole year compared to the summers.

image

Figure 4. Mean summer and mean annual temperatures and monthly standard deviations for historic and future periods.

Download figure to PowerPoint

[14] One of the most critical questions is if the selected years 2001 and 2050 are representative years for both the historic and future period. In order to answer this question comparison for the cumulative distribution function (CDF) plots and spatial distribution plots is conducted for both historic and future periods. The CDF plots for temperature and humidity are similar for the three consecutive historic as well as future years but there is an obvious shift to higher values moving from historic to future period (figures are presented in the auxiliary material). Moreover, the spatial distribution plots show similar trend for the consecutive years in both periods. The CDF's plots for the precipitation are similar between the two periods. The spatial distribution plots for the three consecutive years in both periods have the same pattern with only small local changes.

3.3. Regional Air Quality

[15] Air quality model performance is evaluated by comparing the observed and predicted daily maximum eight (8) hour O3 (M8hO3) and hourly PM2.5 concentrations over the United States (Figures 5 and 6) using data from more than 1000 stations for ozone and about 100 for PM2.5 (see http://www.epa.gov/ttn/airs/airsaqs/detaildata/downloadaqsdata.htm). Around 250 ozone monitoring stations located at west, Midwest and southeast subregions, 200 at northeast and 150 at Plains. Regarding PM2.5 monitoring stations there are around 45 at Plains, 35 at west and northeast and 25 at Midwest and southeast. The three simulated summer mean M8hO3 concentrations for 2000–2002 are about 15% higher, while the PM2.5 concentrations are about 30% lower than the observed. Model performance for the PM2.5 concentrations is significantly more region-dependent than the M8hO3 concentrations. Representation of secondary organic aerosol (SOA) formation is uncertain, and low organic carbon (OC) has been noted in the CMAQ approaches [e.g., Chen and Griffin, 2005; Kroll et al., 2006; Lim and Ziemann, 2005]. Recent work suggests that this is due to lower yields and higher vapor pressures in CMAQ [Morris et al., 2005]. Moreover, the current chemical mechanism neglects isoprene as a SOA precursor, though its role in SOA formation might be quite important [e.g., Claeys et al., 2004a, 2004b; Henze and Seinfeld, 2006], leading to discrepancies between the predicted and observed PM2.5 concentrations. The effect of NOX on SOA yields, which is highly uncertain, has also been neglected.

image

Figure 5. Mean summer (2000–2002) and mean annual (2001) observed and predicted maximum eight (8) hour O3 (M8hO3) concentrations and monthly standard deviations.

Download figure to PowerPoint

image

Figure 6. Mean summer (2000–2002) and mean annual (2001) observed and predicted PM2.5 concentrations and monthly standard deviations.

Download figure to PowerPoint

[16] Annual mean M8hO3 and PM2.5 concentrations are better simulated compared to the three-summer average. Mean annual M8hO3 concentration is slightly (10%) overpredicted. Simulated PM2.5 concentrations are low during spring and summer and high during the rest of the year largely because of the underprediction of organic carbon. The presented standard deviation is calculated for the monthly average concentrations. Ozone concentrations are high during summers and low during the rest of the months resulting in higher annual standard deviation compared to summers in both observations and predictions. The variation in PM2.5 concentration during a year is less than ozone as PM2.5 high concentrations exist during autumn and winter.

3.3.1. Summer Pollutant Changes

[17] Global climate change, alone, has a small effect on future summer (i.e., 2049–2051) M8hO3 concentrations over the United States (Figure 7) when compared to the historic summers (i.e., 2000–2002). The average regional changes range from −2.5% to +2.8% (Table 1a). As noted, Leung and Gustafson [2005] found a small increase in stagnation events, and this, in part, leads to increases in the number of days where concentrations are over 85 ppb in most regions expect the Midwest (Table 1b). Stagnation events are predicted to have the most impact in the west, northeast and Plains and a small impact in the southeast. Summer PM2.5 concentrations (Figure 8) are predicted to be lower in all the U.S. subregions (average about 10%), using the same emission inventory, as a result of the increased precipitation and higher temperatures in spite of higher biogenic VOC emissions. The effect of climate change alone in summer PM2.5 concentrations seems to be quite important in the Midwest, southeast and Plains (Table 1a). Higher temperatures lead to increased gas phase partitioning of ammonium nitrate and organics. Sulfate, nitrate, ammonium and organic carbon decrease because of increased precipitation and higher temperatures (Table 1a) but no significant modification in PM2.5 composition is predicted (Table 1c).

image

Figure 7. Mean summer and mean annual maximum eight (8) hour O3 (M8hO3) concentrations and monthly standard deviations for historic and future periods.

Download figure to PowerPoint

image

Figure 8. Mean summer and mean annual PM2.5 concentrations and monthly standard deviations for historic and future periods.

Download figure to PowerPoint

Table 1a. Mean Summer and Mean Annual Changes (Percent) in Pollutant Concentrations for Future Periods Compared to Historic Ones
 M8hO3, %PM2.5, %SO4, %NO3, %NH4, %OC, %
SummersSummers_npSummersSummers_npSummersSummers_npSummersSummers_npSummersSummers_npSummersSummers_np
West−11.60.9−15.7−2.0−32.2−3.7−72.8−42.8−33.0−6.9−6.70.7
Plains−15.8−0.1−34.3−12.1−48.7−16.4−46.4−15.2−41.8−14.1−16.2−7.7
Midwest−24.4−2.5−37.1−18.4−52.6−22.4−68.5−24.1−45.7−21.9−19.1−11.7
Northeast−20.22.8−41.2−1.7−56.7−2.2−79.3−28.8−44.5−0.8−25.2−0.4
Southeast−27.90.3−45.2−14.3−60.5−16.5−77.1−37.1−47.9−13.3−27.5−14.8
United States−18.90.0−35.9−9.9−52.6−13.9−65.6−22.6−43.9−12.2−17.2−5.5
 M8hO3, %PM2.5, %SO4, %NO3, %NH4, %OC, %
20502050np20502050np20502050np20502050np20502050np20502050np
West−6.50.2−9.22.9−20.24.8−41.4−17.6−24.9−3.44.08.9
Plains−7.91.4−22.0−0.8−29.25.5−45.3−17.9−31.7−3.2−3.44.7
Midwest−10.5−0.2−22.74.2−22.212.6−48.5−7.7−28.74.2−9.36.6
Northeast−10.0−0.5−28.56.5−37.410.3−45.6−4.3−32.65.9−13.010.7
Southeast−14.82.3−31.4−2.4−41.50.5−54.9−12.4−37.0−1.7−14.9−3.6
United States−9.20.9−23.41.1−30.86.2−47.8−12.4−31.6−0.2−6.44.4
Table 1b. Number of Days per Summer Month and per Grid Cell Where M8hO3 Concentration Is Over 85 ppb
RegionSummers 2000–2002Summers 2049–2051Summers 2049–2051_np
West0.150.010.44
Plains1.210.021.56
Midwest4.520.084.22
Northeast2.180.023.37
Southeast6.780.057.11
United States2.480.032.77
Table 1c. Mean Summer and Mean Annual PM2.5 Composition of Pollutants Concentrations for Historic Period, Future Period and Future Period_np (Historic Emissions and Future Meteorology)
 Historic Summers2001
SO4, %NO3, %NH4, %OC, %EC, %Other, %SO4, %NO3, %NH4, %OC, %EC, %Other, %
West21284951519111040515
Plains471151531930171514222
Midwest443141332327221511223
Northeast452132041631171517317
Southeast501142031234131420316
United States442142031729171418319
 Future Summers2050
SO4, %NO3, %NH4, %OC, %EC, %Other, %SO4, %NO3, %NH4, %OC, %EC, %Other, %
West171654418177846418
Plains371141922727121318228
Midwest341121723428151413129
Northeast331122622628131421222
Southeast36013262232991325222
United States321122622726111322226
 Future Summers_np2050_np
SO4, %NO3, %NH4, %OC, %EC, %Other, %SO4, %NO3, %NH4, %OC, %EC, %Other, %
West211850515199942516
Plains451151532132141515222
Midwest423141442330191512321
Northeast452132041633161418316
Southeast480142031535121520315
United States421132132030151419319

[18] The impact of climate change, growth activity and emissions controls are more pronounced for the PM2.5 concentrations than M8hO3 (Figures 7 and 8). The U.S. summer average concentrations for M8hO3 and PM2.5 are predicted to be lower by about 20% and 35%, respectively. Significant reduction is predicted for sulfate, nitrate and ammonium while a smaller reduction is predicted for organic carbon (Table 1a). Sulfate will be a significantly lower fraction of PM2.5 in the future; nitrate and ammonium will be slightly lower but organic carbon is predicted to be higher (Table 1c). Significant reduction is also estimated for the highest M8hO3 concentrations over all U.S. subregions along with the average concentrations (Table 2). The Midwest is simulated to have the highest peak M8hO3 concentrations in the future as climate change alone has a more significant effect compared to the other U.S. subregions. Better air quality is also estimated for the cities and megacities (Table 2). Significant reduction in the number of days that the M8hO3 concentrations exceed the standard of 85 ppb as well as the peak values are estimated for all the cities examined here. Atlanta in the southeast U.S. subregion will benefit more; no days are estimated for the M8hO3 concentrations above the air quality standards. In general, there is little year-to-year variation in region wide M8hO3 concentrations as well as the number of days that the M8hO3 concentrations exceed the standard of 85 ppb as well as the peak values for the cities examined (Table 2). Spatial distribution plots for mean summer ozone and PM2.5 concentrations show the reduction in the higher concentrations simulated at the east comes from emissions control strategies (Figures 8 and 9 in auxiliary material), though lower concentrations may actually increase. Climate change alone leads to increasing concentrations in all cities. Moreover climate change lengthens the stagnations events in these cities, similar to the regional behavior described previously and more days with M8hO3 concentrations over the air quality standard are predicted in Los Angeles, New York and Houston.

Table 2. Regional and Local (Cities) Predicted Maximum Eight (8) Hour O3 (M8hO3) Concentration Characteristicsa
 M8hO3, ppb
Summers 2000–2002Summers 2049–2051Summers 2049–2051_np
99.750%Number of Days Over 85 ppbPeak Value99.750%Number of Days Over 85 ppbPeak Value99.750%Number of Days Over 85 ppbPeak Value
000102000102Average000102495051495051Average49505149np50np51np49np50np51npAverage49np50np51np
  • a

    The regional value corresponds to the 99.750% of the cumulative distribution function concentrations. The local values correspond to the number of days where M8hO3 concentrations exceed the standard of 0.08 ppm as well as the peak estimated concentration.

West/Los Angeles918887543739431211131187773752410171710594961049410376678375146130139
Plains/Houston10197956845365013211611376777691219131009910910210110247706460139130143
Midwest/Chicago116115116342832311321401447678894521101009712411011513217194427137127165
Northeast/New York11010810432243129119114109717584008383838910711112431395441126124135
Southeast/Atlanta11610811078667874130122102778483010082858111211611672757173149133136
3.3.2. Annual Average Pollutant Changes

[19] A separate comparison between the annual average concentrations of M8hO3 and PM2.5 is performed for the future (i.e., 2050) and historic (i.e., 2001) years. Annual average PM2.5 levels tend to be stable year to year. Comparison of the three consecutive summers reveals only small differences (typically less than 10% for M8hO3 and 15% for PM2.5); inclusion of more consecutive yearly data is not expected to change significantly the results of our analysis as no significant weather modification for the consecutive years is estimated (see: auxiliary material). Further evidence is found in observations. Monitoring stations in large U.S. cities (e.g., Los Angeles, New York, Chicago) show a small variation (about 1 μg/m3, or 5–10%) in annual average PM2.5 levels for the years 2000–2002 (http://www.epa.gov/airtrends). This is similar to the observed trend from 1999–2005 showing a decrease of 7% nationwide. The same trend is observed for M8hO3 concentrations showing a decrease of 8% nationwide.

[20] While much of the analysis concentrates on the higher ozone levels found predominantly in the summer, annual statistics are provided as well because some areas have longer ozone seasons, and there is increasing concern over exposures (human and other) to lower ozone levels [U.S. Environmental Protection Agency, 2006]. As is noted, much greater reductions are found for higher ozone levels and in the ozone CDFs (Figure 12, auxiliary material). Others [e.g., Lefohn et al., 1998], have found that intermediate and lower levels of ozone are not as responsive to controls. Further, emission changes can lead to increases in ozone at night and during photochemically less active periods, as seen by examining the low-concentration tail of the CDF (Figure 12, auxiliary material).

[21] Annual average concentrations for both pollutants (M8hO3 (Figure 7) and PM2.5 (Figure 8)) are predicted to be slightly different over the United States in year 2050 compared to 2001, using the 2001 emission inventory (Table 1a). The sulfate and organic carbon fraction of PM2.5 is predicted to be slightly higher while the nitrate fraction lower (Table 1c). This is caused by the higher VOCs and SO2 emissions in a warmer climate, although the same emission inventory is used. The higher SO2 emissions lead to more H2SO4 formation that quickly reacts with NH3 to form ammonium sulfate. On the other hand, the higher NOX emissions, although leading to formation of HNO3, do not translate in increase in nitrate concentrations since nitrate aerosol formation depends on the availability of NH3 after neutralization of H2SO4. Regional changes in future meteorology (e.g., temperature, precipitation, wind) combined with projected emissions, lead to some regional variation in air quality changes (Tables 1a and 1c). More clouds and precipitation in the southeast increase aqueous oxidation and wet deposition leading to a net slight increase in sulfate and a decrease in organic carbon concentrations compared to the rest of the regions.

[22] Impacts of climate change, activity growth and emissions controls are more pronounced for regional PM2.5 concentrations than M8hO3. The annually average U.S. concentrations for PM2.5 and M8hO3 are predicted to be 23% and 9% lower, respectively in 2050 compared to 2001. Significant reductions are predicted for sulfate (−31%), nitrate (−48%), and ammonium (−32%) fractions, while only a small reduction is predicted for organic carbon (−6%) (Table 1a). Controls on NMVOC emissions from area and point sources that are less stringent than for SO2 and NOX combined with the higher VOC emissions from biogenic sources expected in a warmer future climate are the primary factors. A slight increase in organic carbon in the west is noted because of increase in both primary and secondary organic carbon. Sulfate, nitrate and ammonium fractions of PM2.5 are predicted to be lower in the future compared to historic period while organic carbon will be higher (Table 1c). Recent work suggests that SOA formation from both biogenic [e.g., Pun et al., 2003] and anthropogenic [e.g., de Gouw et al., 2004] are larger than have previously been accounted for in atmospheric chemistry models. Further, work by Volkamer et al. [2006] and Mendoza-Dominguez and Russell [2001] also suggests that primary OC emissions may be larger as well. Such findings provide further evidence that OC will be the dominant fine aerosol species in the future. However, they also show that significant uncertainty remains as to the current and potential future source impacts.

[23] Seasonal variation in M8hO3 concentrations gives higher concentrations during summers at all subregions (Table in auxiliary material). The differences between summer and the rest of the seasons seems to diminish in the future under the impact of both climate change and emissions projection as higher reduction is estimated during summer. Seasonal variation in PM2.5 concentrations gives higher values during winter and autumn and lower during spring and summer at all subregions. Reductions are forecast for the average PM2.5 concentrations over all U.S. subregions (Table 3) although climate change can lead to increases. The Midwest is simulated to have the highest daily average PM2.5 concentrations in the future. Lower PM2.5 concentrations are also forecast for the cities. Reduction in the number of days that the daily average PM2.5 exceeds the standard of 35 μg/m3 as well as the peak values are estimated for all the mega cities examined here except the peak value at Los Angeles, although, again, climate change alone leads to increases. Annual average spatial distribution plots for ozone and PM2.5 concentrations show again the reduction in the higher concentrations simulated at the east comes from emissions control strategies (Figures 10 and 11 in auxiliary material.) Comparison between summers and annual distribution plots confirms that ozone is significant problem during summer especially in the east while PM2.5 is important pollutant all over the year.

Table 3. Regional and Local (Cities) Predicted Daily Average PM2.5 Concentration Characteristicsa
Region/CityPM2.5, μg/m3
200120502050np
Region: 99.750%CityRegion: 99.750%CityRegion: 99.750%City
Number of Days Over 35 μg/m3Peak ValueNumber of Days Over 35 μg/m3Peak ValueNumber of Days Over 35 μg/m3Peak Value
  • a

    The regional value corresponds to 99.750% of the cumulative distribution function concentrations. The local values correspond to the number of days where daily average PM2.5 concentrations exceed the standard of 35 μg/m3 as well as to the peak estimated concentration.

West/Los Angeles21.5540.518.6156.921.5867.8
Plains/Houston33.51448.528.3446.134.61154.0
Midwest/Chicago45.23580.041.52363.648.04464.0
Northeast/New York40.03888.832.11259.342.94079.0
Southeast/Atlanta40.03866.034.01854.341.34163.1

4. Conclusions

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results and Discussion
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

[24] Regional O3 and PM2.5 concentrations for a future period (i.e., summers 2049–2051 and year 2050) are simulated to be lower compared to the historic period (i.e., summers 2000–2002 and year 2001), given the planned controls on precursor emissions, though global warming, alone, does lead to an increase in biogenic emissions. Climate change, alone, with no emissions growth or controls has a small effect on the M8hO3 and PM2.5 levels although changes in stagnation events, leading to higher pollutant concentrations over a slightly extended duration, may be regionally important. Future levels of sulfate, nitrate and ammonium are simulated to be significantly lower compared to organic carbon, leaving organic carbon as the likely major constituent of fine particulate matter in the far future. M8hO3 concentrations over all domain subregions are simulated to be lower than the historic scenarios; both the number of days with M8hO3 concentrations above the standards and the peak concentrations are reduced for the urban areas.

[25] The trend in pollutant concentrations reveals the key role that emission control strategies may play in future regional air quality, setting forecasting of emissions as key to being able to assess the impact of climate change on pollutant concentrations. One of the most important implications of this study is that the significant reduction predicted for sulfate, nitrate and ammonium concentrations will result in organic carbon as the most important PM2.5 component.

[26] These results are going to be further used for studying the sensitivity of future pollutant concentrations to emission changes as well as the uncertainties in regional air quality and changes in sensitivities to climate change uncertainties and source-specific emissions (K.-J. Liao et al., Sensitivities of ozone and fine particulate matter formation to emissions under the impact of potential future climate change, submitted to Environmental Science and Technology, 2007; K.-J. Liao et al., Climate impacts on air quality and response to controls: Not such an uncertain future, manuscript in preparation, 2007).

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results and Discussion
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

[27] This work was supported by U.S. EPA Science To Achieve Results (STAR) grants: RD83096001, RD82897602 and RD83107601. The views expressed in this paper are those of the authors and do not necessarily reflect the views or policies of the EPA. The Pacific Northwest National Laboratory is operated for the U.S. Department of Energy by Battelle Memorial Institute under contract DE-AC06-76RLO 1830. We would like to acknowledge Loretta Mickley from Harvard University for the GISS simulations used by L. R. Leung.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results and Discussion
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information
  • Bergin, M. S., G. S. Noblet, K. Petrini, J. R. Dhieux, J. B. Mildford, and R. A. Harley (1999), Formal uncertainty analysis of a Lagrangian photochemical air pollution model, Environ. Sci. Technol., 33, 11161126.
  • Binkowski, F. S., and S. J. Roselle (2003), Models-3 Community Multiscale Air Quality (CMAQ) model aerosol component: 1. Model description, J. Geophys. Res., 108(D6), 4183, doi:10.1029/2001JD001409.
  • Brasseur, G. P., and E. Roeckner (2005), Impact of improved air quality on the future evolution of climate, Geophys. Res. Lett., 32, L23704, doi:10.1029/2005GL023902.
  • Chen, J., and R. J. Griffin (2005), Modeling secondary organic aerosol formation from oxidation of a-pinene, b-pinene, and d-limonene, Atmos. Environ., 39, 77317744.
  • Claeys, M., W. Wang, A. C. Ion, I. Kourtchev, A. Gelencser, and W. Maenhaut (2004a), Formation of secondary organic aerosols from isoprene and its gas-phase oxidation products through reaction with hydrogen peroxide, Atmos. Environ., 38, 40934098.
  • Claeys, M., et al. (2004b), Formation of secondary organic aerosols through photooxidation of isoprene, Science, 303, 11731176.
  • Cohan, D. S., D. Tian, Y. T. Hu, and A. G. Russell (2006), Control strategy optimization for attainment and exposure mitigation: Case study for ozone in Macon, Georgia, Environ. Manage., 38(3), 451462.
  • de Gouw, J. A., P. D. Goldan, C. Warneke, W. C. Kuster, J. M. Roberts, M. Marchewka, S. B. Bertman, A. A. P. Pszenny, and W. C. Keene (2004), Validation of proton transfer reaction-mass spectrometry (PTR-MS) measurements of gas-phase organic compounds in the atmosphere during the New England Air Quality Study (NEAQS) in 2002, J. Geophys. Res., 109, D22301, doi:10.1029/2004JD004690.
  • El-Fadel, M., and M. Massoud (2000), Particulate matter in urban areas: Health-based economic assessment, Sci. Total Environ., 257, 133146.
  • Galizia, A., and P. L. Kinney (1999), Long-term residence in areas of high ozone: Associations with respiratory health in a nationwide sample of nonsmoking young adults, Environ. Health Perspect., 107(8), 675679.
  • Grell, G., J. Dudhia, and D. R. Stauffer (1994), A description of the fifth generation Penn State/NCAR mesoscale model (MM5), NCAR Tech. Note, NCAR/TN-398+STR,Natl. Cent for Atmos. Res., Boulder, Colo.
  • Henze, D. K., and J. H. Seinfeld (2006), Global secondary organic aerosol from isoprene oxidation, Geophys. Res. Lett., 33, L09812, doi:10.1029/2006GL025976.
  • Hogrefe, C., B. Lynn, K. Civerolo, J.-Y. Ku, J. Rosenthal, C. Rosenzweig, R. Goldberg, S. Gaffin, K. Knowlton, and P. L. Kinney (2004), Simulating changes in regional air pollution over the eastern United States due to changes in global and regional climate and emissions, J. Geophys. Res., 109, D22301, doi:10.1029/2004JD004690.
  • Intergovernmental Panel on Climate Change (1996), Climate Change 1995: The Scientific Basis, Cambridge Univ. Press,, Cambridge, U. K.
  • Intergovernmental Panel on Climate Change (2000), Emissions Scenarios, Cambridge Univ. Press, Cambridge, U. K.
  • Intergovernmental Panel on Climate Change (2001), Climate Change 2001: The Scientific Basis, Cambridge Univ. Press, Cambridge, U. K.
  • Kleinman, L. I. (2000), Ozone process insights from field experiments—part II: Observation based analysis of ozone production, Atmos. Environ., 34, 20232034.
  • Knowlton, K., J. E. Rosenthal, C. Hogrefe, B. Lynn, S. Gaffin, R. Goldberg, C. Rosenzweig, K. Civerolo, J. Y. Ku, and P. L. Kinney (2004), Assessing ozone related health impacts under a changing climate, Environ. Health Perspect., 112(115), 15571563.
  • Kroll, J. H., N. L. Ng, S. M. Murphy, R. C. Flagan, and J. H. Seinfeld (2006), Secondary organic aerosol formation from isoprene photooxidation, Environ. Sci. Technol., 40, 18691877.
  • Langner, J., R. Bergstrom, and V. Foltescu (2005), Impact of climate change on surface ozone and deposition of sulphur and nitrogen in Europe, Atmos. Environ., 39, 11291141.
  • Lefohn, A. S., D. S. Shadwick, and S. D. Zim (1998), The difficult challenge of attaining EPA's new ozone standard, Environ. Sci. Technol., 32(11), 276A282A.
  • Leung, L. R., and W. I. Gustafson Jr. (2005), Potential regional climate and implications to U.S. air quality, Geophys. Res. Lett., 32, L16711, doi:10.1029/2005GL022911.
  • Lim, Y. B., and P. J. Ziemann (2005), Secondary organic aerosol formation from reactions of n-alkanes with OH radicals in the presence of NOx, Environ. Sci. Technol., 39, 92299236.
  • Mendoza-Dominguez, A., and A. G. Russell (2001), Estimation of emission adjustments from the application of four-dimensional data assimilation to photochemical air quality modeling, Atmos. Environ., 35, 28792894.
  • Mickley, L. J., D. J. Jacobs, B. D. Field, and D. Rind (2004), Effects of future climate change on regional air pollution episodes in the United States, Geophys. Res. Lett., 31, L24103, doi:10.1029/2004GL021216.
  • Morris, R. E., D. E. McNally, T. W. Tesche, G. Tonnesen, J. W. Boylan, and P. Brewer (2005), Preliminary evaluation of the community multiscale air, quality model for 2002 over the southeastern United States, J. Air Waste Manage. Assoc., 55(11), 16941708.
  • Murazaki, K., and P. Hess (2006), How does climate change contribute to surface ozone change over the United States? J. Geophys. Res., 111, D05301, doi:10.1029/2005JD005873.
  • National Research Council (2001), Global Air Quality: An Imperative for Long-Term Observational Strategies, 41 pp., Natl. Acad. Press, Washington, D. C.
  • Pekkanen, J., K. L. Timonen, J. Puuskanen, A. Reponen, and A. Mirme (1997), Effects of ultrafine and fine particles in urban air on peak expiratory flow among children with asthmatic symptoms, Environ. Res., 74, 2433.
  • Pun, B. K., S. Y. Wu, C. Seigneur, J. H. Seinfeld, R. J. Griffin, and S. N. Pandis (2003), Uncertainties in modeling secondary organic aerosols: Three-dimensional modeling studies in Nashville/western Tennessee, Environ. Sci Technol., 37(16), 36473661.
  • Rind, D., J. Lerner, K. Shah, and R. Suozzo (1999), Use of on line tracers as a diagnostic tool in general circulation model development: 2. Transport between the troposphere and the stratosphere, J. Geophys. Res., 104, 91239139.
  • Russell, A. G., and R. Dennis (2000), Photochemical air quality modeling: NARSTO critical review, Atmos. Environ., 34, 22832324.
  • Ryerson, T. B., et al. (2001), Observations of ozone formation in power plant plumes and implications for ozone control strategies, Science, 292, 719723.
  • Schwartz, J., P. Michaels, and R. E. Davis (2005), Ozone: Unrealistic scenarios, Environ, Health Perspect., 113(2), A86A87.
  • U.S. Environmental Protection Agency (2006), Air quality criteria for ozone and related photochemical oxidants, EPA 600/R-05/004aF, Washington, D. C.
  • Volkamer, R., J. L. Jimenez, F. San Martini, K. Dzepina, Q. Zhang, D. Salcedo, L. T. Molina, D. R. Worsnop, and M. J. Molina (2006), Secondary organic aerosol formation from anthropogenic air pollution: Rapid and higher than expected, Geophys. Res. Lett., 33, L17811, doi:10.1029/2006GL026899.
  • Webster, P. J., G. J. Holland, L. A. Curry, and H. R. Chang (2005), Changes in tropical cyclone number, duration and intensity in a warming environment, Science, 309, 18441846.
  • Woo, J. H., S. He, P. Amar, E. Tagaris, K. Manomaiphiboon, K. J. Liao, and A. G. Russell (2006), Development of mid-century anthropogenic emissions inventory in support of regional air quality modeling under influence of climate change, paper presented at 15th Annual Emission Inventory Conference “Reinventing Inventories—New Ideas in New Orleans,”, U.S. Environ. Prot. Agency, New Orleans, La., 16–18 May. (Available at http://www.epa.gov/ttn/chief/conference/ei15/session4/woo2.pdf).
  • Zhang, Y., P. Liua, B. Punb, and C. Seigneur (2005), A comprehensive performance evaluation of MM5-CMAQ for the summer 1999 Southern Oxidants Study episode—Part I: Evaluation protocols, databases, and meteorological predictions, Atmos. Environ., 40, 48254838.

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results and Discussion
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

Auxiliary material for this article contains 12 figures and 1 table.

Auxiliary material files may require downloading to a local drive depending on platform, browser, configuration, and size. To open auxiliary materials in a browser, click on the label. To download, Right-click and select “Save Target As…” (PC) or CTRL-click and select “Download Link to Disk” (Mac).

See Plugins for a list of applications and supported file formats.

Additional file information is provided in the readme.txt.

FilenameFormatSizeDescription
jgrd13692-sup-0001-readme.txtplain text document4Kreadme.txt
jgrd13692-sup-0002-fs01.epsPS document7971KFigure S1. Regional CDF plots for temperature in 2000: comparison between observations and predictions.
jgrd13692-sup-0003-fs02.epsPS document7959KFigure S2. Regional CDF plots for temperature in 2001: comparison between observations and predictions.
jgrd13692-sup-0004-fs03.epsPS document7974KFigure S3. Regional CDF plots for temperature in 2002: comparison between observations and predictions.
jgrd13692-sup-0005-fs04.epsPS document10145KFigure S4. CDF plots for temperature, humidity and precipitation over US.
jgrd13692-sup-0006-fs05.epsPS document2481KFigure S5. Spatial distribution plots of the annual average temperature.
jgrd13692-sup-0007-fs06.epsPS document2469KFigure S6. Spatial distribution plots of the annual average humidity.
jgrd13692-sup-0008-fs07.epsPS document2961KFigure S7. Spatial distribution plots of the annual average precipitation.
jgrd13692-sup-0009-fs08.epsPS document1992KFigure S8. Spatial distribution plots of the three summer average ozone concentrations in historic years (O3_2000-2002 summers) and future years (O3_2049-2051 summers).
jgrd13692-sup-0010-fs09.epsPS document1732KFigure S9. Spatial distribution plots of the three summer average PM2.5 concentrations in historic years (PM2.5_2000-2002 summers) and future years (PM2.5_2049-2051 summers).
jgrd13692-sup-0011-fs10.epsPS document2984KFigure S10. Spatial distribution plots of the annual average ozone concentrations in 2001 (O3_2001) and 2050 (O3_2050).
jgrd13692-sup-0012-fs11.epsPS document2325KFigure S11. Spatial distribution plots of the annual average PM2.5 concentrations in 2001 (PM2.5_2001) and 2050 (PM2.5_2050).
jgrd13692-sup-0013-fs12.epsPS document10707KFigure S12. Daily maximum 8 hour ozone concentration CDF plots in 2001, 2050 and 2050_np (2001 emission inventory and 2050 meteorology).
jgrd13692-sup-0014-ts01.txtplain text document1KTable S1. Seasonal M8hO3 and PM2.5 concentrations over US sub-regions in years 2001, 2050 and 2050_np (2001 emissions inventory and 2050 meteorology).
jgrd13692-sup-0015-ts01.xlsapplication/excel14KTable S1. Seasonal M8hO3 and PM2.5 concentrations over US sub-regions in years 2001, 2050 and 2050_np (2001 emissions inventory and 2050 meteorology).
jgrd13692-sup-0016-t01.txtplain text document1KTab-delimited Table 1a.
jgrd13692-sup-0017-t02.txtplain text document0KTab-delimited Table 1b.
jgrd13692-sup-0018-t03.txtplain text document1KTab-delimited Table 1c.
jgrd13692-sup-0019-t04.txtplain text document1KTab-delimited Table 2.
jgrd13692-sup-0020-t05.txtplain text document1KTab-delimited Table 3.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.