Projections of Future Summertime Ozone over the U.S

©2014 American Geophysical Union. All rights reserved.

Projections of Future Summertime Ozone over the U.S.

G. G. Pfister1, S. Walters1, J.-F. Lamarque1, J. Fast2, M. C. Barth1,3, J. Wong1,4, J. Done3, G. Holland3, C. L. Bruyère3,5

1 Atmospheric Chemistry Division, NCAR, Boulder, CO 2 Pacific Northwest National Laboratory, Richland, WA 3 Mesoscale and Microscale Meteorology, NCAR, Boulder, CO 4 University of Colorado, Boulder, CO 5 Environmental Sciences and Management, North-West University, South Africa

This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1002/2013JD020932


Abstract

This study uses a regional fully coupled chemistry-transport model to assess changes in

surface ozone over the summertime U.S. between present and a 2050 future time period at

high spatial resolution (12 km grid spacing) under the SRES A2 climate and RCP8.5

anthropogenic precursor emission scenario. The impact of predicted changes in regional

climate and globally enhanced ozone is estimated to increase surface ozone over most of the

U.S; the 5th - 95th percentile range for daily 8-hour maximum surface ozone increases from

31-79 ppb to 30-87 ppb between the present and future time periods. The analysis of a set of

meteorological drivers suggests that these mostly will add to increasing ozone, but the set of

simulations conducted is targeted on understanding the competing effect of global versus

local emission changes and does not allow separating the meteorological feedbacks from that

due to enhanced global ozone. Statistically significant increases were found for future

temperature, biogenic emissions and solar radiation. Stringent emission controls can

counteract these likely positive feedbacks and if implemented as in RCP8.5, we estimate large

reductions in surface ozone with the 5th-95th percentile reduced to 27-55 ppb.

A comparison of the high-resolution projections to global model projections shows that the

global model has a high positive bias in surface ozone compared to the regional model and

compared to observations. On average, both the global and the regional model predict similar

changes in ozone between the present and future time periods. However, on small spatial

scales, the regional model shows pronounced differences between projections in urban and

rural regimes that cannot be resolved at the coarse (2°x2°) resolution of the considered global

model.

This study confirms the key role of emission control strategies in future air quality

projections and demonstrates the need for considering degradation of air quality with future

climate change in emission policy making. It also illustrates the need for high resolution

modeling when the objective is to address regional and local air quality or establish links to

human health and society.


1. Introduction

Modern society, commerce and industry are becoming increasingly complex and

vulnerable to changes in weather, air quality and climate, emphasizing the need for reliable

projections of future climate and air quality on regional to local scales. Trace gases and

aerosols in the atmosphere affect climate, both in direct and indirect ways. The direct

influence is through greenhouse gases such as methane, ozone, nitrous oxide or the chloro-

fluoro-carbons absorbing infrared radiation and releasing heat, as well as through various

aerosols absorbing and scattering solar radiation. Atmospheric chemical processes also

directly impact atmospheric levels of carbon dioxide, through oxidation of hydrocarbons that

produce carbon dioxide. The indirect influence occurs through the modification of cloud

properties (e.g. aerosols acting as cloud condensation nuclei) affecting water resources,

deposition of absorbing black carbon on snow surfaces, and various interactions with the

biosphere (e.g. acid rain, ozone damage to vegetation, and nitrogen deposition).

In turn, regional air quality is also strongly influenced by weather and climate. Recent

studies [Camalier et al., 2007; Bloomer et al., 2009] have highlighted the strong correlation

between surface ozone concentrations and surface temperature. Camalier et al. [2007] find

that as much as 80% of maximum daily 8-hour ozone in the eastern U.S. can be explained by

a generalized linear model with the two most important predictors temperature (positive) and

relative humidity (negative). This dependence is partly due to increased reactivity at higher

temperatures or potentially more frequent stagnation events, but also arises from increased

local emissions of ozone precursors, in particular biogenic emissions, which are strongly and

nonlinearly temperature dependent. In a future climate where extreme weather events may

become more frequent, heat waves and drought conditions may also increase the likelihood of

wildfires with adverse effects on air quality and human health. The interplay between

emission and weather change and relevant processes such as convection, topographical

effects, city pollution, etc. cannot be explicitly resolved within a global model and to assess

the sensitivity of air quality to changes in climate requires high-resolution modeling.

This paper focuses on changes in surface ozone over the contiguous U.S. as a result of

future changes in emissions and/or climate. We perform a set of climate simulations with

chemistry at high-resolution (12 km grid spacing) to predict changes in summertime surface

ozone between present time and a future time period centered around 2050. The 2050 time

period was chosen as a compromise between having a sufficient climate signal while at the

same time have some level of confidence in the future emission projections. Time periods in

the upcoming decades are also more relevant to air quality managers and decision makers.

Understanding the way surface ozone is impacted by changes in climate and emissions is of


major importance for human health, society and economy, but the direction of the change is

not always clear because of multiple competing factors.

Previous studies [Hogrefe et al., 2004; Dentener et al., 2006; Steiner et al., 2006; Nolte et

al., 2008] have demonstrated the potential increase in future ozone levels under a warmer

climate while at the same time a decrease in global background ozone is predicted due to

enhanced loss rates as a result of increased absolute humidity [Racherla et al., 2006; Liao et

al., 2006; Murazaki and Hess, 2006; Wu et al., 2008a, Doherty et al., 2013]. Predictions of

changes in the global tropospheric ozone burden, however, depend on whether effects of

climate on natural ozone sources are accounted for. Increased lightning-generated and soil

NOx emissions, biogenic non-methane volatile organic compounds (NMVOCs) or enhanced

stratospheric-tropospheric exchange likely are leading to an increase in future global

tropospheric ozone [Zeng and Pyle, 2003; Sanderson et al., 2003; Hauglustaine et al., 2005;

Wu et al., 2008b, Chen et al., 2009; Hegglin and Shepherd, 2009; Kawase et al., 2011;

Lamarque et al., 2011; Young et al., 2013]. Furthermore, increased methane levels contribute

to growing global background ozone [Fiore et al., 2002; West and Fiore, 2005; West et al.,

2006; Young et al., 2013].

Over the U.S., potentially large decreases in surface ozone are simulated when reductions

in precursor emission estimates are accounted for, counteracting potential increases due to

climate change [e.g. Tagaris et al., 2007; Tao et al., 2007; Zhang et al., 2008; Lam et al.,

2011]. For example, Nolte et al. [2008] estimated reductions of summertime ozone over the

U.S. of ~15 ppb by 2050 under the IPCC A1B scenario [IPCC, 2000]. Similar reductions are

found [Kelly et al., 2012] using the Supplemental Report on Emission Scenarios (SRES) A2

climate scenario and emissions from the IPCC Representative Concentration Pathway (RCP)

6 scenario.

Most previous studies of future air quality projections were conducted at either global

scales and fairly coarse spatial resolution [Prather et al., 2003; Mickley et al., 2004; Coquard

et al., 2004; Dentener et al., 2006; Murazaki and Hess, 2006], or at higher resolution but for a

limited number of years [Hogrefe et al., 2004; Nolte et al., 2008; Zhang et al., 2008a; Weaver

et al., 2009; Avise et al., 2012]. Few studies have used regional offline models over multiple

years to project air quality changes under future climate conditions and most of them focused

on Europe [Meleux et al., 2007; Katragkou et al., 2011; Huszar et al., 2012, Juda-Rezler et

al., 2012]. Over the U.S., Kelly et al. [2012] used a regional model at 45 km resolution to

conduct 10 years of present and future air quality simulations.

Given the large inter-annual variability in weather and climate, simulations spanning

multiple years are needed to derive robust statistics of future changes, while at the same time

high spatial resolution is vital when addressing air quality. This study distinguishes from


previous work in that we conduct simulations of multiple years (13 summers) for each present

and future air quality scenario using a fully coupled regional chemical transport model at high

spatial resolution (12 km grid spacing) over the contiguous U.S. Simulations are initialized

each April and results for June-August are analyzed. The simulations follow the RCP 8.5

anthropogenic emission scenario [vanVuuren et al., 2011], and the SRES A2 climate scenario.

The regional climate projections are based on global simulations from the Community

Climate System Model (CCSM 3.0), which were dynamically downscaled to a 36 x 36 km2

domain over the larger North-America region [Done et al., 2013]. The latter provided input

for meteorological initial (IC) and boundary conditions (BC) for the regional chemistry-

climate simulations. Chemical IC and BC are based on chemistry climate simulations from

the global Community Atmosphere Model with Chemistry (CAM-Chem). At the time the

global simulations were performed, global climate simulations for RCP 8.5 were not available

and for this reason CAM-Chem in the same way as the regional simulations, follows the

SRES A2 climate and RCP8.5 emission scenario.. The SRES A2 and RCP 8.5 are high

climate impact scenarios with 8 W m-2 and 8.5 W m-2 radiative forcing by 2100, respectively.

Given that both scenarios agree closely in their overall global radiative forcing, it can be

expected that the respective climate projections will also be close and within the range of

natural climate variability and inter-model variations.

The objective of this study is to provide an estimate of how surface ozone over the

summertime U.S. is expected to change for assumed changes in climate and two different

U.S. emission scenarios: one with a significant reduction in ozone precursor emissions and

one where emissions are held at present time levels. The paper is organized as follows. In

Section 2 we describe the model setup and inputs, followed by the evaluation of modeled

surface ozone with in-situ measurements in Section 3. The results are discussed in Section 4

looking at predicted changes in surface ozone over the U.S. (Section 4.1), presenting an

analysis of the meteorological and chemical drivers behind these changes (Section 4.2) and

providing a comparison between regional and global projections of surface ozone (Section

4.3). The paper closes with a summary (Section 5).


2. Simulations

Global climate simulations provide the basic input to drive the regional chemistry climate

simulations. The global climate projections are based on global simulations from the

Community Climate System Model (CCSM 3.0), which were dynamically downscaled to a 36

x 36 km2 domain over the larger North-America region [Done et al., 2013]. The latter

provided input for meteorological initial (IC) and boundary conditions (BC) for the regional

chemistry-climate simulations. Present time and future chemical IC and BC were taken from

a simulation with the global Community Atmosphere Model with Chemistry (CAM-Chem)

[Lamarque et al., 2011] for a comparable climate and anthropogenic emission scenario.

2.1. NRCM-Chem Code and Setup

The nested regional climate model (NRCM) with chemistry (NRCM-Chem) is based on

the regional Weather and Forecasting model (WRF) with Chemistry (WRF-Chem, version

3.3), which is a fully coupled chemical transport model [Grell et al., 2005]. NRCM-Chem has

been updated to include climate relevant processes and feedbacks such as aerosol direct and

indirect effects, dry and wet deposition of gases and aerosols, and changes in stratospheric

ozone, while at the same time keeping computational costs manageable. The major features

that were added include a reduced hydrocarbon chemical scheme (REDHC) [Houweling et

al., 1998] linked to physics parameterizations ported from the global Community Atmosphere

Model version 5 (CAM5) including the Modal Aerosol Model (MAM) [Liu et al., 2012], a

cloud macrophysics scheme similar to [S. Park, C.S. Bretherton, P.J. Rasch, Cloud

Simulation in the Community Atmosphere Model, 5; to be submitted to J. Clim.] , a two-

moment cloud microphysics scheme [Morrison and Gettelman, 2008], a shallow convection

scheme [Park and Bretherton, 2009], and a moist boundary layer scheme [Bretherton and

Park, 2009]. The implementation of the CAM5 physics into WRF-Chem is described and

evaluated in Ma et al. [2013].

The configuration was also adapted to work with existing model processes such as online

biogenic emissions, which is especially important for future simulations given their strong

dependence on solar radiation and temperature, and are calculated following the Model of

Emissions of Gases and Aerosols from Nature (MEGAN) [Guenther et al., 2006]. Land use

and land cover are kept constant between present and future period and, as such, changes in

biogenic emissions are entirely due to changes in the meteorological drivers. Lightning as a

source of NOx in the upper troposphere was not considered for either the present or future

simulations, but lightning-NOx emissions have been shown to not significantly affect overall

levels of surface ozone [Kaynak et al., 2008]. Other model configurations include the Noah

land surface model [Chen and Dudhia, 2001], the Grell 3D Ensemble convective scheme and


the improved Rapid Radiative Transfer Model (RRTMG) for representing short- and long-

wave radiative transfer through the atmosphere [Iacono et al., 2008].

The NRCM-Chem domain covers the wider contiguous U.S. (Figure 1) at 12 km grid

spacing in a Mercator projection [20N to 54N, -140 to -51W]. There are 697 x 349 horizontal

grid cells and 51 vertical levels from the surface up to 10 hPa. The meteorological time step is

60 seconds and the chemical time step 120 seconds. The model is not run as one continuous

simulation, but is initialized each April and run through September; the focus of our analysis

is on the three-month period from June through August (JJA) allowing for a 2-month spin-up

phase. Each set of simulations is run for 13 summers (1996-2008) and includes a present time

base simulation (SIMPRES), a future time simulation centered around 2050 (SIMFUT; 2046-

2058) with changed climate and future anthropogenic emission projections, and a second

future time simulation (SIMFUT_CLIM), which simulates future climate and global emissions,

i.e. has future initial and boundary conditions, but over the regional domain uses present time

anthropogenic precursor emissions (Table 1). Instantaneous output of relevant 3D

meteorological and chemical fields is written every 3 hours, while output is written every

hour for essential surface parameters including surface ozone and other air quality relevant

species such as nitrogen oxides and aerosols, 2m temperature, 10m winds, precipitation and

solar radiation.

2.2. Meteorological Initial and Boundary Conditions

The NRCM-Chem simulations are driven by meteorological initial conditions (IC) and 6-

hourly boundary conditions (BC) downscaled from a NRCM 36 km x 36 km simulation

described in [Done et al., 2013]. Here we provide just a brief overview of this simulation. The

NRCM is also based on the WRF model [Skamarock and Klemp, 2008], but uses an earlier

version and different configuration than NRCM-Chem. NRCM is nested into a global climate

model with options selected for long-term simulation, as described by Leung and Gustafson

[2005] and Done et al. [2011]. The NRCM is free to generate its own weather and climate

constrained only by atmospheric data at the lateral boundaries and sea surface temperatures at

the lower boundary. Global climate data are provided by CCSM3.0 [Collins et al., 2006]

integrations run under the SRES A2 scenario (IPCC SRES SPM 2001), as submitted to the

Coupled Model Intercomparison Project 3 [Meehl et al., 2007a]. The CCSM is a full Earth

system model, including atmosphere, ocean, cryosphere, biosphere, and land surface. These

data are bias corrected as described by Bruyere et al. [2013] and then used to drive the

NRCM 36 km domain using one-way nesting.

The NRCM 36km domain simulations are for 10 years for both time periods [1996-2005

and 2046-2055] and provide IC and BC for 10 years of NRCM-Chem. For the additional

three years of NRCM-Chem simulations the initial start date from NRCM output was shifted


by one week. These extension years allow for more robust statistics and also can be used to

assess the sensitivity of climate simulations to initial conditions (not discussed here). During

the NRCM-Chem setup sea surface temperatures (SST) were inadvertently held constant in

the NRCM-Chem simulations thereby missing seasonality or future changes as was accounted

for in NRCM. While this will lead to differences in the NRCM-Chem climate compared to

the NRCM climate, this impact is mitigated by having a model domain that largely is situated

over land and by reinitializing the NRCM-Chem simulations each April. We carefully

evaluated the meteorological variables between the two models, and the differences we find

over land are in a range that could be explained by just all other differences between the

NRCM and NRCM-Chem setup such as different model physics, resolution or chemistry

feedbacks. Caution will need to be taken in analyzing coastal areas where SSTs have a

stronger influence on the meteorology. Table 2 gives a brief summary on the comparison

between NRCM and NRCM-Chem and lists statistics of different meteorological parameters

over the 10 years both NRCM and NRCM-Chem simulations are available for. NRCM-Chem

has somewhat lower surface temperature and as a result a drier atmosphere (lower water

vapor and less rainfall), but overall is within the range of values calculated for NRCM and

agrees closely in the predicted future changes; e.g. a nearly 2C increase in the mean surface

temperature over the U.S. and an increase in JJA total rainfall by on average 10 mm. NRCM-

Chem also follows well the year-to-year variability of the different meteorological variables

as is simulated by NRCM (not shown here).

2.3. Chemical Initial and Boundary Conditions

Chemical IC and BC for trace gases and aerosols are derived from a global simulation

with the Community Atmosphere Model with Chemistry (CAM-Chem V4). The global

simulations consider the effects of both changes in climate and anthropogenic emissions and

do not allow separation of the individual impacts on the initial and boundary concentrations

that are fed into the regional model. Details describing these simulations are found in

Lamarque et al. (2011). As no climate simulations following RCP8.5 were available at the

time of the CAM-chem simulations, previously performed climate simulations following the

SRES A2 scenario [Meehl et al., 2007b] were used to define the sea-surface temperature and

sea-ice distributions. The 2000-2100 CAM-chem simulation was performed with tropospheric

and stratospheric chemistry on a 1.9ox2.5o grid and a 30 minute timestep, starting from initial

conditions generated from an equivalent 1850-2000 simulation [Lamarque et al., 2010]. The

tropospheric gas phase scheme is identical to the one included in NRCM-Chem, but CAM-

Chem uses a different lumped aerosol scheme, which is based on work by Tie et al. [2005].

Output of the necessary reactive species and aerosols is available as monthly averages

from CAM-Chem and these fields are linearly interpolated in time to the 6-hour NRCM-


Chem BCs. Each present time NRCM-Chem year uses the same chemical IC and BC based

on the CAM-Chem output for the year 2000. Similarly, each future NRCM-Chem year uses

CAM-Chem simulations for the year 2050 for IC and BC. While this does not account for any

possible year-to-year variability in long-range transport in NRCM-Chem, it allows us to better

separate future changes in long-range transport and background chemical composition from

other driving factors (e.g. emission changes). Another reason for this is that the meteorology

between the CAM-Chem and NRCM-Chem years is not the same; therefore inter-annual BCs

will not match between the simulations. Analysis of the global model fields also shows that

inflow at the lateral boundaries varies little between years, at least in terms of monthly

averages (not shown here). The concentrations of methane, H2 and N2O are provided through

the initial conditions and held constant at this level throughout the simulation period. The

average surface methane concentrations for the present and future time period are 1.8 ppm

and 2.8 ppm, respectively. This increase in methane is a strong driver of the increase in global

levels of ozone [Lamarque et al., 2011] and will also impact ozone production over the U.S..

The WRF-Chem model does not include stratospheric chemistry, but typically has

stratospheric species controlled by IC and BC. Especially for long-term simulations, however,

an adequate representation of the stratosphere and its future changes are crucial. Here we use

a scheme similar to what is included in the NCAR global models MOZART-4 [Emmons et

al., 2010] and CAM-Chem [Lamarque et al., 2012] in which upper chemical boundary

conditions are imposed to keep key species (e.g. ozone, N2O, HNO3,CO, methane,...)

representative of stratospheric levels. Concentrations of these species are set to a

climatological value above 50 hPa and between this level and the tropopause relaxed to the

climatology. The climatological fields for the stratosphere are generated from monthly-

averaged output of the global CAM-chem simulations, in which there is a representation of

stratospheric chemistry, including heterogeneous ozone depletion [Lamarque et al., 2008;

Lamarque and Solomon, 2010].

2.4. Emissions

Anthropogenic emission inputs for present and future time periods at high resolution were

developed by regridding the global RCP emissions [Lamarque et al., 2011] included in the

CAM-Chem simulations from 0.5° to 0.1° using the finer 0.1° spatial information from

EDGAR-4.1 (http://edgar.jrc.ec.europa.eu). Emissions for the present time period are for the

year 2000 and we selected the RCP8.5 anthropogenic emission scenario (most extreme

climate scenario with a 8 W m-2 radiative forcing by 2100) for the future simulations. The

anthropogenic emissions are emitted at the lowest model level and no diurnal cycle was

applied because this information is not available from the global inventories. Figure 1 shows

the spatial distribution of anthropogenic NOx emissions for JJA over the NRCM-Chem


domain. The downscaled resolution allows individual urban areas to show up as well as major

highways. Present time NOx emission totals for the modeling domain for JJA are 2.35 Tg N

and these are reduced to about half in the future scenario over the contiguous U.S. (Table 2).

Reductions are due to control strategies implemented to U.S. and Canadian sources, while the

growth of the industrial sector in Mexico leads to increased emissions. Ship emissions are

also estimated to slightly increase by the 2050’s (Fig. 1).

For fire emissions we created an annual hourly climatology based on fire emissions for

2002-2010 from the NCAR Fire model FINN [Wiedinmyer et al., 2011]. These emissions are

different from what is used in the global CAM-Chem model, which uses biomass burning

emissions from the RCP 8.5 scenario. However, the latter also show only small changes over

the U.S. between the present and future scenario. FINN provides fire emissions for gas and

aerosol species at 1 km x 1 km on a global scale based on satellite derived fire counts. This

climatology is used for both present and future time periods. We chose to implement the same

fire emissions in the future scenarios that are used in the present–day simulation in order to

reduce the number of drivers affecting surface ozone. Another reason is the high uncertainty

in predictions of future fire emissions. The fire emissions in the model are changing daily and

hourly with a diurnal profile following recommendations by the Western Regional Air

Partnership (WRAP; Report to Project No. 178-6, July 2005). We apply the online plume rise

module within NRCM-Chem [Freitas et al., 2007] to estimate the emissions injection height

and vertically distribute fire emissions. The average fire NOx emissions for JJA are included

in Figure 1. Total NOx fire emissions over the domain for the 3-month period amount to 46

GgN; emission totals for other species are listed in Table 2. As stated in Section 2.1, the

emissions of NMVOC from vegetation are calculated online in WRF-Chem using the

MEGAN model, which is also used in CAM-Chem. Biogenic emissions are influenced by

temperature and solar radiation, but we do not account for the effects of changing landcover

or changing CO2 concentrations [Guenther et al., 2006]; the latter could lead to a slight

overestimate in the increase in biogenic emissions under a warmer future climate [Heald et

al., 2009].

3. Statistical Evaluation of Present Time Surface Ozone

To evaluate the representativeness of the model in simulating summertime surface ozone

over the U.S. we compare simulated ozone to hourly ozone measurements from the EPA Air

Quality System (AQS; http://www.epa.gov/ttn/airs/airsaqs/detaildata/downloadaqsdata.htm).

Since the model simulates climate conditions and does not reflect the specific time periods of

the observations, the evaluation is done on a statistical basis. We use observations for 1996-

2008 over the U.S. and select monitoring sites with at least five years of data for JJA, which

yields about 1200 monitoring sites included in the comparison. Modeled hourly surface ozone


from the 1998-2010 simulation years is extracted for each grid cell that contains a monitoring

site, i.e. the model is interpolated in space but not in time to the observations.

Figure 2a shows the observed and modeled diurnal cycle of surface ozone averaged for

JJA over the contiguous U.S. The NRCM-Chem results follow well the observed diurnal

cycle and also capture the observed variability. The spatial correlation ranges from ~0.5

during nighttime to 0.8 during local afternoon. Modeled surface ozone shows a high bias of

about +5 ppb during daytime and a low bias of similar magnitude during nighttime compared

to the observations. The average ozone mixing ratio for 14-18 LT (0-4 LT) is 57±21ppb

(17±16 ppb) in the model compared to 51±21 ppb (22±16 ppb) in the observations.

Uncertainties such as the boundary layer height evolution might be contributing to the

diurnally varying bias, but also the omission of a diurnal cycle in anthropogenic emissions

can add to the differences. The model further uses a year 2000 specific inventory for all

simulation years and as such does not reflect changes in emissions occurring during the 1996

to 2008 observational period the observations are referring to.

As indicted by the good correlation and shown in more detail in Figure 2b, the model

provides a good representation of the spatial distribution of surface ozone for local afternoon

and nighttime with highest daytime ozone in central California and the eastern U.S. of 70 ppb

or more. The overall high daytime bias in the model is also evident in the spatial distribution

and is to a large degree attributed to an overestimate of eastern U.S. surface ozone, where the

majority of the monitoring sites are located. Most models produce too much ozone over the

eastern U.S. but the reasons are not fully understood [Fiore et al., 2009; Yu et al., 2010;

Rasmussen et al., 2012]. NRCM-Chem reflects well the prominent ozone gradient across sites

in California and low ozone values along the coastal regions in the southern U.S. The model

also overall captures well the nighttime surface ozone distribution, and reflects elevated ozone

concentrations at individual sites in Virginia, North Carolina, the north-eastern corner of the

U.S., and high elevation sites in the western U.S.

Figure 2c presents cumulative and bin-size frequency distributions for observed and

modeled surface ozone for local afternoon and nighttime. The shift in the model values to

slightly higher daytime and slightly lower nighttime values is evident, however, the

distribution shape and day-to-night changes are well captured. At nighttime the model shows

a significantly larger contribution of very low values, which is partially due to the model not

representing the nighttime reduction in NOx emissions, but also indicates general issues of

models in simulating nighttime chemistry. What further has to be kept in mind is that the

detection limit of the surface monitors is typically around 3-5 ppb [Colorado Department of

Public Health and Environment, personal communication].


Given that the WRF-Chem model has been widely used and evaluated in previous studies

[Frost et al., 2006; Kim et al., 2009; Zhang et al., 2010; Jiang et al., 2012; Barth et al., 2012]

and that the model captures the main features in surface ozone over different climatic regions

during the present time period simulation, it is reasonable to assume that the model can be

used to examine future changes in surface ozone concentrations and distributions.

4. Future Changes in Surface Ozone

We first quantify changes between present and future ozone from the NRCM-Chem

simulations, evaluate how concentrations and air quality related metrics are expected to

change, then examine some of the driving factors behind future changes. This is followed by a

discussion that compares future ozone projections between NRCM-Chem and CAM-Chem to

assess the impact of different modeling systems and spatial resolutions

4.1. Regional Projections of Surface Ozone

Figure 3 shows the spatial distribution of average surface ozone for SIMPRES, SIMFUT and

SIMFUT_CLIM and their differences. The results are shown for the average maximum daily 8-

hour average surface ozone (MDA8) during JJA. The significance of the results is evaluated

with a Student’s T-Test for a 95% confidence level with lower confidence levels masked in

the difference plots. As can be seen, the 13 individual summer simulations allow for robust

results across most of the contiguous U.S. In support of this analysis, we show in Figure 4 the

frequency distributions of MDA8 for the different simulations (limited to grid cells between

28-50N over the contiguous U.S.).

The combined effects from changing climate and reducing emissions (Figure 3a, b and c)

leads to a strong decrease in surface ozone over the contiguous U.S. with the strongest

absolute reductions (up to ~40 ppb) in areas of high surface ozone and high anthropogenic

precursors emissions such as the eastern U.S. and California. The average MDA8 over the

U.S. decreases from 53±15 ppb to 40±8 ppb and the range spanned by the 5th and 95th

percentile is reduced from 31-79 ppb to 27-55 ppb between present and future time periods

(Figure 4). Our estimates are about the order of the 20% (11-28%) reduction in mean summer

MDA8 over the U.S found by Tagaris et al. [2007] for the combined effect of emissions and

climate, who applied a less extreme scenario (A1B scenario). Higher reductions, changes in

MDA8 exceeding 15 ppb, were found by Nolte et al. [2008] also using the A1B scenario,

however that study did not consider the impact of rising methane levels on the ozone

background concentrations.

The large reduction in ozone over the contiguous U.S. also leads to a reduction in the

outflow of ozone from the East Coast to the Atlantic Ocean (Figure 3b). At the West Coast

we see a sharp contrast between the negative ozone change over land and a positive change in


surface ozone of up to ~10 ppb over the Pacific Ocean. The latter is due to increased inflow at

the lateral boundaries. Higher ozone at the boundaries is a result of enhanced global ozone

production mostly due to increased levels of methane [Lamarque et al., 2011]. To a lesser

extent the increased ship emissions also contribute to this ozone increase. Changes in

synoptic-scale transport patterns might further explain some of the differences, but they are

not examined here.

When emissions are kept at present time levels in a future climate (SIMFUT_CLIM),

simulated surface ozone over most of the domain increases by up to ~10 ppb (Figure 3e and

d) with the largest absolute changes in California, over the northeastern U.S. and also the

Rocky Mountain region (Figure 3d). Mean MDA8 over the U.S. is estimated to increase by

about 4 ppb over SIMPRES and the range spanned by the 5th and 95th percentile increases to 30-

87 ppb (Figure 4). These changes reflect the impact of a changing regional climate and

enhanced ozone entering the regional domain through the lateral boundaries and also in an

increase in ozone production over the U.S. through increased levels of methane.

Hogrefe et al. [2004] using a 36 x 36 km2 air quality simulation for five consecutive

summers estimated that regional climate change alone will increase summertime MDA8 over

just the eastern U.S. by a similar amount. They, however, conclude that given the strong

interannual variability, simulations for more than five years would be needed. Lam et al.

[2011] also studied the effect of just climate change for the less extreme A1B climate scenario

using three years of regional simulations over the Eastern half of the U.S. and report an

increase in MDA8 of 2-2.5 ppb.

Figure 3f demonstrates the impact when reducing emissions under a future climate

scenario (SIMFUT-SIMFUT_CLIM). We find a similar pattern as when changing emissions and

climate (Figure 3b), but reduction in ozone levels over the U.S. are more pronounced given

that the climate effects now do not dampen the effects of emission reductions. Figure 4

demonstrates that the distribution of MDA8 is rather similar between SIMFUT and

SIMFUT_CLIM, but the latter is shifted to higher values.

The results are also presented in a different way in Figure 5 using as a metric the average

change in the number of exceedance days per summer. An exceedance day is defined as a day

with a maximum 8-hour ozone concentration larger than 75 ppb, which conforms to the

current ozone National Air Quality Standard (NAAQS), and for maximum 8-hour ozone

concentration larger than 65 ppb, which reflects a possible lowering of the NAAQS in the

near future. For SIMFUT the number of exceedance days is significantly reduced, but the

model projections suggest that despite stringent emissions controls some areas over the U.S.

can still exceed current standards. For example, ozone over the high elevation areas of the

Rocky Mountains regions is more strongly impacted by background concentrations and long-


range transport, both of which are expected to increase in a future scenario. Also stratospheric

intrusions can impact surface levels over this region, specifically during spring and early

summer and might become more prevalent in a future climate [Collins et al., 2003; Murazaki

and Hess, 2006; Zeng and Pyle, 2003]. In addition, stratospheric air reaching the surface

might be more ozone rich given that the amount of stratospheric ozone is expected to increase

in the future due to the recovery of the stratospheric ozone layer [Eyring et al., 2010]. Future

increase in stratospheric ozone levels is taken into account in NRCM-Chem through the upper

and lateral boundary conditions (Section 2.3).

Without emission controls, the number of high ozone events show a clear increase over

most of the contiguous U.S. We find the number of cases of MDA8 > 75 ppb to increase by

about 70% and cases with MDA8 > 65 ppb to increase by about 50%. In comparison, for the

simulations where emissions controls are assumed, the number of cases with MDA8 > 75 ppb

is less than 1% of current projections (less than 2% for MDA8 > 65 ppb).

4.2. Regional Projections in meteorological and chemical drivers

Relevant climate variables that affect air pollution, biogenic emissions, and deposition

include temperature, relative humidity, solar radiation, clouds, precipitation, boundary layer

heights and wind speed and direction. As shown in the previous Section, without changing

emissions over the U.S., surface ozone is predicted to increase by the 2050’s as a result of

both increased background (inflow) of ozone and as a result of the changing climate. In this

Section we now discuss how some of the meteorological drivers are predicted to change in a

future climate. Given that ozone concentrations can be impacted on time scales shorter than

the summertime averages discussed here, this analysis can only provide a first order estimate

of the meteorological impact on surface ozone. With a limited set of simulations it is also not

possible to detangle the effects of the different contributing factors or assess the impact each

change has on the overall change in surface ozone. Results from previous studies, however,

can provide underlying information in putting the changes in the climatic variables in our

simulation into the context of changes in surface ozone. Figure 6 illustrates present time

conditions and predicted future changes (SIMPRES-SIMFUT) of JJA mean daytime 2m

temperature, specific humidity, downward solar radiation, biogenic isoprene emissions, and

boundary layer height. Changes in these variables are shown for the 95% significance level.

The according statistics over the contiguous U.S. are listed in Table 4.

Temperature is predicted to increase on average by +1.7K over the domain in response to

increased levels of greenhouse gases by 2050; the largest increase of ~3K is seen over the

western U.S. (Figure 6a). Similar characteristics were also reported by other studies [Leung

and Gustafson, 2005; Murazaki and Hess, 2006; Tagaris et al., 2007; Weaver et al., 2009],

but the distribution and degree of warming varies as a result of different climate scenario,


model configuration, and resolution. Studies such as Aw and Kleeman [2003] show that

summertime ozone increases as temperature increases and that this sensitivity is largely

driven by a decrease in the net formation of peroxyacetyl nitrate [Sillman and Samson, 1995].

Thus, the increase in surface temperature in our simulations should exacerbate ozone

pollution throughout the entire model domain. Estimates from previous studies over the

Eastern U.S. report an increase of 0.34 ppbV in regional ozone for a 1K increase in

temperature [Bell et al., 2007; Nolte et al., 2008; Dawson et al., 2008]. These studies are

typically based on the analysis of present day correlations, which might not necessarily be

representative under future climate conditions. For this reason, fully interactive models are

needed to project how future changes in climate variables impact the chemical composition in

the troposphere.

A warmer atmosphere has the capacity to hold more water vapor suggesting an increase in

the amount of atmospheric water vapor. This is illustrated in Figure 6b, which shows a

significant increase in specific humidity of up to ~10% over the model domain under the

future climate scenario. Largest absolute changes occur over the Central and Eastern U.S.

Comparable values were found by Murazaki and Hess [2006], who for 2100 estimate an

increase in surface temperature of 2-4K and in water vapor by 10-20%. While increased

water vapor on a global scale leads to enhanced ozone loss rates in the troposphere [Stevenson

et al., 2006], at the surface, elevated water vapor is estimated to increase ozone in regions of

high anthropogenic emissions, while it may cause a slight decrease in more remote NOx

limited regimes [Steiner et al., 2006; Jacob and Winner, 2009].

Downward shortwave radiation shown in Figure 6c increases over the contiguous U.S., but

there are also large parts over the contiguous U.S. where due to interannual variability no

significant change is detected. Increased solar radiation indicates a decrease in cloudiness,

which would lead to enhanced ozone production. It has to be considered that simulated

changes in cloudiness strongly depend on the choice of the microphysical schemes and can

vary widely amongst different schemes and modeling systems. Higher downward shortwave

radiation together with higher temperatures will lead to enhanced biogenic NMVOC

emissions and increased availability of ozone precursors, specifically in the southern U.S.,

California and parts of the intermountain western U.S. (Figure 6d). The emissions of isoprene

over the domain for JJA are predicted to increase from 16.4 Tg C to 18.7 Tg C between

SIMPRES and SIMFUT. In our study we did not account for changes in land use and previous

studies indicate a smaller change in isoprene emissions with predicted changes in land use

compared to when using a fixed vegetation map [Sanderson et al., 2003; Wu et al., 2012] but

at the same time also point towards the large uncertainty in future predictions of land use and

land cover. Increased VOCs might lead to a slight decrease in surface ozone in areas of low


NOx concentrations, while in high NOx areas surface ozone is estimated to increase with

increased biogenic VOCs [Steiner et al., 2006; Wu et al., 2012].

Mixing and transport within the boundary layer are important mechanisms for diluting and

transporting pollutants. The predicted increase in boundary layer height in the intermountain

west will act to enhance ventilation and reduce high pollution buildup, while a change in the

opposite direction is predicted in other areas (Figure 6e). The average change in boundary

layer height is highly variable, and similar to downward shortwave radiation shows large

areas within the U.S. with no statistical significant change over time. Wind speed is even

more variable and over the U.S. shows very little to no significant change between the present

and future scenario (not shown here).

Two more integrative metrics to examine the potential for pollution buildup and that

combine various meteorological conditions are the air stagnation index (ASI) or the number

of unvented hours. Similar to Horton et al. [2012] we calculate a stagnant day when daily

mean 500 hPa wind speed is less than 13 m/s, daily mean 10-m wind speed is less than 3.2

m/s, and daily total precipitation is below 1 mm; with the caveat that NRCM-Chem output for

upper-level wind speed is only available every six hours and hence does not reflect changes in

between this time interval. The number of unvented hours is determined by counting the total

time when the ventilation coefficient, i.e. the product of boundary layer height and mean wind

speed within the boundary layer is below 6000 m2/s [Leung and Gustafson, 2005]. Figure 7

shows for both metrics maps of present time distribution as well as predicted future changes.

While the two different metrics deviate on smaller spatial scales, the overall patterns are

similar. Stagnant conditions in a current climate are most common along the west coast and

over the southeastern U.S., while the central U.S. is less prone to stagnation in part, due to the

mountainous terrain. Under future meteorological conditions we find that stagnant conditions

might increase over large parts of the U.S., particularly in the eastern half of the U.S., which

indicates an overall increased potential for pollution buildup. In contrast, increased ventilation

and an overall reduction in ASI is predicted over parts of the intermountain west, consistent

with the above discussion on increased boundary layer. Nevertheless, increased solar

radiation, temperature, and biogenic emissions might counteract some of the possible ozone

reductions resulting from dynamic processes. Mostly we find that changes in stagnant

conditions and the mean changes in wind speed and PBLH have a similar spatial pattern, but

given that ASI and the ventilation coefficient are determined by thresholds and not by mean

values, it is possible that areas with increased stagnation show an increase in mean wind and

PBLH or vice versa.

While it is not possible to separate the impacts of globally enhanced ozone and regional

climate on future projections of surface ozone from the set of simulations conducted, the


analysis of the above set of meteorological drivers suggests that these to a large degree will

act to increase surface ozone. Statistically significant changes are found for temperature,

water vapor, and biogenic emissions, which are also assumed to be stronger drivers behind

overall ozone changes compared to solar radiation, boundary layer height and wind speed,

which show large regions where no statistical change can be detected.

4.3. Regional and Global Model Projections

Higher resolution models can provide a major advantage over coarse resolution global

models in predicting future surface ozone by providing more detailed information on local

and regional changes. In addition to this, ozone production in a coarse grid is likely

overestimated because the emissions of ozone precursors are artificially diluted [Kim et al.,

2010; Mayer et al.,2000; Liang and Jacobson, 2000; Sillman et al., 1990,], which at the same

time can cause an overestimate in nighttime ozone especially in urban regimes due to reduced

titration. In this Section we compare surface ozone projections from the global CAM-Chem

model with our NRCM projections to assess the impacts of model resolution. Even though

model resolution is a major difference, CAM-Chem and NRCM-Chem also differ in physics

(CAM-4 versus CAM-5 physics), dynamics and various inputs, e.g. fire emissions or

lightning NOx; the latter is considered in CAM-Chem but not in NRCM-Chem (Sections 2.3

and 2.4). As such the comparison provides a more general assessment of air quality

projections as derived from different modeling systems. Since CAM-Chem results are

available only for 10-year periods, we also only include 10 years of the NRCM-Chem

simulations in this analysis.

Figure 8 compares the spatial distribution of mean JJA daytime (16-22 UTC) surface

ozone for present and future time periods from CAM-Chem and NRCM-Chem and also

shows the respective surface ozone frequency distributions over the contiguous U.S. (defined

by land areas within 28-50N). The overall spatial variations and distribution functions from

both CAM-Chem and NRCM-Chem are similar, but the global model simulates about +10

ppb higher ozone values compared to NRCM-Chem. A similar bias between the global and

regional projections is also found for the future time period, and as a result the two models

agree closely in the range of their predicted changes in surface ozone. This suggests that the

main drivers of ozone change occur on scales that can be resolved by a global model, at least

for the considered future scenario. As expected, the global model is not able to resolve small-

scale patterns, but on a larger scale both models give similar projections: the largest

reductions are found in the eastern U.S. and strong reductions are also seen around isolated

metropolitan areas in Colorado and Arizona.

The analysis for nighttime (6-10 UTC) shown in Figure 9 provides similar conclusions,

but shows a more pronounced influence of model grid resolution. Nighttime ozone in NRCM-


Chem is expected to decrease over most of the domain, except in cities where the

combination of reduced emissions and reduced titration causes nighttime ozone to increase.

CAM-Chem predicts increased ozone over some of the larger urban areas in the Eastern U.S.,

Florida and California, but for most cities the global projections are opposite in sign from

NRCM-Chem.

Both models predict increased daytime (Figure 8) and nighttime (Figure 9) ozone in the

inflow region over the Pacific Ocean and reduced ozone outflow over the Atlantic in line with

what has been discussed in Section 4.1. Along the U.S. West Coast the two models shows

more pronounced differences in the spatial pattern and the sign of ozone change. A coarse

model resolution might merge grid boxes over land and ocean and cannot separate the sharp

contrast in emissions, land type characteristics and chemical regime between land and ocean.

The global model resolution will also not be able to simulate the complex terrain of coastal

and mountainous terrain in this region. The good agreement of NRCM-Chem simulations

with measurements from surface monitors (Figure 2) for California and the representation of

large urban metropolitan areas such as the Bay Area or L.A., give confidence that the regional

model is capable of representing these areas, but given the omission of varying SST in

NRCM-Chem, the results of the regional model for coastal areas also might need to be

interpreted with caution.

At the East coast both models project reduced outflow in the future as a result of reduced

ozone levels, but the outflow region in NRCM-Chem is more widespread compared to CAM-

Chem, specifically during nighttime. This is in parts due to differences in simulated ozone

concentrations over the contiguous U.S. but also reflects differences in simulated dynamics

and physics, which are not analyzed in here.

4.4. Resolving Regional and Local Scale with High Resolution Modeling

Finally, we compare CAM-Chem and NRCM-Chem over a subset of the regional domain

to look at local and regional impacts in more detail and how these are resolved at different

model resolutions. We focus on the Intermountain West, which is a high elevation area

subject to complex topography and isolated metropolitan areas. The major cities in the region

include Denver (Colorado), Salt Lake City (Utah),, Phoenix (Arizona),, and Albuquerque

(New Mexico). Similar to before we show in Figures 10 and 11 projected changes in surface

ozone from NRCM-Chem and CAM-Chem for daytime and nighttime, respectively, but now

focus on the selected region only.

NRCM-Chem shows elevated daytime surface ozone (Figure 10) near urban areas, for

which the locations are indicated in the maps, as well as their impacts on the surrounding

area. These are also the areas where the largest future reductions in surface ozone are

estimated given that these are strong source regions and most impacted by reduced emissions.


The global CAM-Chem simulation, in contrast, simulates high surface ozone over most of the

region and does not resolve individual urban areas. Reductions in future ozone for the area

around Denver do stand out over the background reductions and there is a larger signal

around Salt Lake City, but clearly at coarse resolution the localized impacts cannot be

resolved. As before CAM-Chem shows a similar shape of the frequency distribution function

for surface ozone, but is shifted higher by about 10 ppb in absolute concentrations. The

frequency distributions are similar for future changes, but CAM-Chem shows a higher

number of high negative ozone changes.

The nighttime ozone distribution in NRCM-Chem (Figure 11) reflects reduced nighttime

ozone in areas with high emissions such as the aforementioned cities, as well as over other

source regions like the Four-Corner power plants at the border of Utah, Colorado, New

Mexico and Arizona and economic areas in the Texas panhandle. Ozone is now higher in

remote areas where it is not or not significantly depleted due to titration. High ozone is

specifically found over the Rocky Mountain region where to some degree free tropospheric

air can reach down to the surface. Periods of upslope winds can also transport high daytime

ozone from the urban sources to higher elevation areas. The CAM-Chem simulation does not

reflect this recirculation flow well.

In the future scenario, nighttime ozone in the NRCM-Chem simulation is predicted to

decrease over most of the domain by up to 10 ppb except in emission hotspots where the

decreased NOx emissions cause ozone to increase by up to about the same amount. CAM-

Chem simulates reductions of comparable magnitude, but only predicts ozone increase over

the largest source regions (Denver and Salt Lake City) while all other areas show a decrease.

These results demonstrate that in a more general sense both a regional and the global model

provide similar conclusions in regard to future air quality projections but that coarse

resolution models clearly are limited in providing region-specific information or information

relevant to address air quality specific issues.

5. Summary

Poor air quality is a result of high emissions and unfavorable weather conditions. Emission

controls enforced over the U.S. have led to improvements in air pollution in the recent

decades, but there is concern that efforts in controlling emissions could be offset by future

changes in climate, i.e. future changes in weather conditions. Given that air quality varies on

local scales, high spatial resolution modeling is needed to investigate the links between

emissions, chemistry and weather and provide future projections of climate and air quality.

This study uses a regional fully coupled chemistry-transport model (NRCM-Chem) to

assess changes in surface ozone over the summertime U.S. between the present and a 2050


future time period under the A2 climate scenario. The simulations are rather unique in that

they combine a large number of consecutive years (13 simulation years per simulation period)

with high spatial resolution (12 km grid spacing). We present a first analysis of this

comprehensive simulation data set with a focus on surface ozone. Modeled surface ozone

over the U.S. during the present time is evaluated by comparison to hourly ozone

measurements from the EPA air quality monitoring network.

Projections of future air quality are strongly dependent on the assumptions that are made

for future climate and emission scenarios, which itself are subject to large uncertainties, as

well as on the modeling setup and on the type of feedbacks considered (e.g. ozone production

from increased methane or effects of future land-use change). As a result, future projections

might show conflicting results in their specific outcomes. Common to most studies and

confirmed by our work, however, is the finding of the key role of emission control strategies

in future air quality projections and a need for considering degradation of air quality with

future climate change in emission policy making.

Our study finds that the impact of predicted changes in climate and globally enhanced

ozone will lead to an increase in surface ozone by 2050 over most of the U.S. with the 5th-95th

percentile range for daily 8-hour maximum surface ozone increasing from 31-79 ppb in

present time to 30-87 ppb. This increase also is associated with an increased number of cases

when the daily maximum 8-hour surface ozone exceeds current and potential future national

standards. It is not possible from the set of simulations conducted to separate the impacts of

globally enhanced ozone, which is mostly a result of a 1 ppm increase in methane by 2050

and has a strong impact on the regional ozone change, from the impacts the increased

methane and regional climate have on net ozone production over the U.S., However, the

projected changes in meteorological drivers, such as a future increase in temperature,

biogenic emissions and solar radiation, suggest that these mostly will also add to increasing

ozone.

However, if stringent emission controls are considered such as from the RCP8.5 scenario,

significant reductions in surface ozone are projected surpassing the positive feedback of

globally increased ozone and regional climate. In this case, the 5th-95th percentile range is

reduced to 27-55 ppb. While the number of cases of daily maximum 8-hour surface ozone

greater than the national standards is going down considerably, areas impacted by background

and transported ozone and possibly stratospheric ozone still show the occurrence of high

ozone events. Overall, the number of cases when the number of daily maximum 8-hour

surface ozone exceeds the current 75 ppb standards is predicted to be less than 1% compared

to the present time period. This is compared to an increase by about 70% if only the impacts


of future changes in regional climate and global ozone are considered, but emissions are kept

at present levels.

Most previous studies of future ozone projections applied global models at coarse

resolution, which aside from not resolving small-scale features tend to overestimate daytime

surface ozone due to dilution. A comparison of our NRCM-Chem projections to projections

from the global CAM-Chem model that is used for initial and chemical boundary conditions

allowed, despite differences in the model configurations, to evaluate the models against each

other and gain a first-order estimate of the impact of model resolution. While the global

projections of surface ozone concentrations are clearly high compared to the regional

estimates and also too high compared to observations, both the global and the regional model

agree rather well in the overall predicted changes between future and present time ozone. This

suggest that the climate effect and anthropogenic emission changes as assumed by the RCP

8.5 scenario, which are similar between the global and the regional model, are the strongest

drivers of future ozone changes and that these feedbacks are resolved at coarser model

resolution. Fire emissions or lightning NOx, which are treated differently in the models, show

a smaller impact.

On smaller spatial scales, though, the regional projections give stronger signals, showing

some regions giving a different sign in the change of ozone, and resolve differences between

urban and rural environments, which are smeared out in the global model. The results

demonstrate that care has to be taken when predicting future air quality changes, not only on

the type of assumptions made in regard to climate and emissions, but also on how models are

being configured. The model spatial resolution is also important and different resolutions can

lead to different and in parts conflicting answers in certain regions and for certain conditions.

This is an important consideration when output from a coarse resolution model is used to

project changes in air quality relevant metrics such as daily 8-hour maximum ozone or

exceedance statistics.

Acknowledgements

We like to acknowledge the three reviewers for their valuable and constructive input. This

work was supported by NSF-EaSM grant AGS-1048829 "Developing a Next-Generation

Approach to Regional Climate Prediction at High Resolution”. Model simulations were made

possible by an NCAR Accelerated Scientific Discovery (ASD) grant. The CESM project

(which includes CAM-chem) is supported by the National Science Foundation and the Office

of Science (BER) of the U. S. Department of Energy. NCAR is operated by the University

Corporation of Atmospheric Research under sponsorship of the National Science Foundation.

Jerome Fast was supported by the DOE’s Office of Science/ Biological and Environmental


Research, through the Atmospheric System Research program at Pacific Northwest National

Laboraotry (PNNL). PNNL is operated for DOE by Battelle Memorial Institute under contract

DE-AC06-76RLO 1830.

References

Avise, J., R.G. Abraham, S.H. Chung, J. Chen, B. Lamb, E.P. Salathe, Y. Zhang, C.G. Knolte, D.H. Loughlin, A. Guenther, C. Wiedinmyer, T, Duhl (2012), Evaluating the effects of climate change on summertime ozone using a relative response factor approach for policymakers, Journal of the Air and Waste Management Society, 62, 1061-1074.

Aw, J., and M. Kleeman (2003), Evaluating the first-order effect of intraannual temperature variability on urban air pollution, Journal of Geophysical Research-Atmospheres, 108(D12), doi:10.1029/2002JD002688.

Barth, M., J. Lee, A. Hodzic, G. Pfister, W. Skamarock, J. Worden, J. Wong, and D. Noone (2012), Thunderstorms and upper troposphere chemistry during the early stages of the 2006 North American Monsoon, Atmospheric Chemistry and Physics, 12(22), 11003-11026, doi:10.5194/acp-12-11003-2012.

Bell, M. L., Goldberg, R., Hogrefe, C., Kinney, P. L., Knowlton, K., Lynn, B., Rosenthal, J., Rosenzweig, C., and Patz, J. A. (2007), Climate change, ambient ozone, and health in 50 US cities, Clim. Change, 82, 61–76.

Bloomer, B., J. Stehr, C. Piety, R. Salawitch, and R. Dickerson (2009), Observed relationships of ozone air pollution with temperature and emissions, Geophysical Research

Letters, 36, doi:10.1029/2009GL037308. Bretherton, C. S., and S., Park, 2009, A new moist turbulence parameterization in the

Community Atmosphere Model, Journal of Climate, 22, 3422-3448,doi:10.1175/2008jcli2556.1.

Bruyere, C., J. Done, G. Holland, S. Fredrick (2013), Bias corrections of global models for regional simulations of high-impact weather, Climate Dynamics, doi: 10.1007/s00382-013-2011-6.

Camalier, L., W. Cox, and P. Dolwick (2007), The effects of meteorology on ozone in urban areas and their use in assessing ozone trends, Atmospheric Environment, 41(33), 7127-7137, doi:10.1016/j.atmosenv.2007.04.061.

Chen, F., and J. Dudhia (2001), Coupling an advanced land surface-hydrology model with the Penn State-NCAR MM5 modeling system. Part I: Model implementation and sensitivity, Monthly Weather Review, 129(4), 569-585, doi:10.1175/1520-0493(2001).

Chen, J., J. Avise, A. Guenther, C. Wiedinmyer, E. Salathe, R.B. JAckson, B. Lamb (2009), Future land use and land cover influences on regional biogenic emissions and air quality in the United States, Atmospheric Environment, 43, 5771-5780.

Collins, W., et al. (2006), The Community Climate System Model version 3 (CCSM3), Journal of Climate, 19(11), 2122-2143, doi:10.1175/JCLI3761.1.

Collins, W., R. Derwent, B. Garnier, C. Johnson, M. Sanderson, and D. Stevenson (2003), Effect of stratosphere-troposphere exchange on the future tropospheric ozone trend, Journal

of Geophysical Research-Atmospheres, 108(D12), doi:10.1029/2002JD002617. Coquard, J., P. Duffy, K. Taylor, and J. Iorio (2004), Present and future surface climate in the

western USA as simulated by 15 global climate models, Climate Dynamics, 23(5), 455-472, doi:10.1007/s00382-004-0437-6.

Dawson, J. P., Racherla, P. N., Lynn, B. H., Adams, P. J., and Pandis, S. N. (2008), Simulating present-day and future air quality as climate changes: model evaluation, Atmos. Environ., 42, 4551–4566.

Dentener, F., et al. (2006), The global atmospheric environment for the next generation, Environmental Science & Technology, 40(11), 3586-3594, doi:10.1021/es0523845.


Doherty, R. M., et al. (2013), Impacts of climate change on surface ozone and intercontinental ozone pollution: A multi-model study, J. Geophys. Res. Atmos., 118, doi:10.1002/jgrd.50266.

Done, J.M., G.M. Holland, C.L. Bruyere, L.R. Leung, A. Suzuki-Parker (2013), Modeling high impact weather and climate: Lessons from a tropical cyclone perspective, Climatic

Change, doi:10.1007/s10584-013-0954-6. Emmons, L., et al. (2010), Description and evaluation of the Model for Ozone and Related

chemical Tracers, version 4 (MOZART-4), Geoscientific Model Development, 3(1), 43-67. Eyring, V. et al. (2010), Multi-model assessment of stratospheric ozone return dates and

ozone recovery in CCMVal-2 models, Atmospheric Chemistry Physics, 10, 9451–9472, doi:10.5194/acp-10-9451-2010.

Fiore, A., et al. (2009), Multimodel estimates of intercontinental source-receptor relationships for ozone pollution, Journal of Geophysical Research-Atmospheres, 114, doi:10.1029/2008JD010816.

Fiore, A., D. Jacob, B. Field, D. Streets, S. Fernandes, and C. Jang (2002), Linking ozone pollution and climate change: The case for controlling methane, Geophysical Research

Letters, 29(19), doi:10.1029/2002GL015601. Freitas, S., K. Longo, R. Chatfield, D. Latham, M. Dias, M. Andreae, E. Prins, J. Santos, R.

Gielow, and J. Carvalho (2007), Including the sub-grid scale plume rise of vegetation fires in low resolution atmospheric transport models, Atmospheric Chemistry and Physics, 7(13), 3385-3398, doi:10.5194/acp-7-3385-2007.

Frost, G., et al. (2006), Effects of changing power plant NO(x) emissions on ozone in the eastern United States: Proof of concept, Journal of Geophysical Research-Atmospheres, 111(D12), doi:10.1029/2005JD006354.

Grell, G., S. Peckham, R. Schmitz, S. McKeen, G. Frost, W. Skamarock, and B. Eder (2005), Fully coupled "online" chemistry within the WRF model, Atmospheric Environment, 39(37), 6957-6975, doi:10.1016/j.atmosenv.2005.04.027.

Guenther, A., T. Karl, P. Harley, C. Wiedinmyer, P. Palmer, and C. Geron (2006), Estimates of global terrestrial isoprene emissions using MEGAN (Model of Emissions of Gases and Aerosols from Nature), Atmospheric Chemistry and Physics, 6, 3181-3210.

Hauglustaine, D., J. Lathiere, S. Szopa, and G. Folberth (2005), Future tropospheric ozone simulated with a climate-chemistry-biosphere model, Geophysical Research Letters, 32(24), doi:10.1029/2005GL024031.

Heald, C.L., M.J. Wilkinson, R.K. Monson, C.A. Alo, G. Wang, A. Guenther (2009), Response of isoprene emissions to ambient CO2 changes and implications for global budgets, Glob. Change Biol., 15:1127-1140, doi:10.1111//j.1365-2486.2008.01802.x

Hegglin, M. I. and T.G. Shepherd (2009), Large climate-induced changes in ultraviolet index and stratosphere-to-troposphere ozone flux, Nature Geosci., 2, 687, doi:10.1038/ngeo604.

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. Horton, D., Harshvardhan, and N. Diffenbaugh (2012), Response of air stagnation frequency

to anthropogenically enhanced radiative forcing, Environmental Research Letters, 7(4), doi:10.1088/1748-9326/7/4/044034.

Houweling, S., F. Dentener, and J. Lelieveld (1998), The impact of nonmethane hydrocarbon compounds on tropospheric photochemistry, Journal of Geophysical Research-

Atmospheres, 103(D9), 10673-10696, doi:10.1029/97JD03582. Huszar, P., K. Juda-Rezler, T. Halenka, H. Chervenkov, D. Syrakov, B.C. Krueger, P. Zanis,

D. Melas, E. Katragkou, M. Reizer, W. Trapp, M. Belda (2011), Effects of climate change on ozone and particulate matter over Central and Eastern Europe, Clim. Res., 50, 51-68.


IPCC, (2000), Emission Scenarios, Nebojsa Nakicenovic and Rob Swart (Eds.), Cambridge University Press, UK. pp 570

Iacono, M. J., J. S. Delamere, E. J. Mlawer, M. W. Shephard, S. A. Clough, and W. D. Collins (2008), Radiative forcing by long-lived greenhouse gases: Calculations with the AER radiative transfer models, J. Geophys. Res., 113, D13103, doi:10.1029/2008JD009944.

Jacob, D. J., and D. A. Winner (2009), Effect of climate change on air quality, Atmos. Environ., 43, 51–63, doi:10.1016/j.atmosenv.2008.09.051.

Jiang, X., C. Wiedinmyer, and A. Carlton (2012), Aerosols from Fires: An Examination of the Effects on Ozone Photochemistry in the Western United States, Environmental Science &

Technology, 46(21), 11878-11886, doi:10.1021/es301541k. Juda-Rezler, K., M. Reizer, P. Huszar, B.C. Krueger, P. Zanis, D. Syrakov, E. Katragkou, W.

Trapp, D. Melas, H. Chervenkov, I. Tegoulias, T. Halenka (2012), Modelling the effects of climate change on air quality over Central and Eastern Europe: concept, evaluation and projections, Clim. Res., 53, 179-203.

Katragkou, E., P. Zanis, I. Kioutsioukis, T. Tegoulias, D. Melas, B.C. Krueger, E. Coppola (2011), Future climate change impacts on summer surface ozone from regional climate-air quality simulations over Europe, J. Geophys. Res., 116, D22307, doi:10.1029/2011JD015899.

Kawase, H., T. Nagashima, K. Sudo, K. Nozawa (2011), Future changes in tropospheric ozone under Representative Concentration Pathways (RCPs), Geophys. Res. Lett., 38, L05801, doi:10.1029/2010GL046402.

Kaynak, B., Y. Hu, R. V. Martin, A. G. Russell, Y. Choi, Y. Wang, (2008) The effect of lightning NOx production on surface ozone in the continental United States, Atmos. Chem.

Phys., 8, 5151-5159, doi:10.5194/acp-8-5151-2008. Kelly, J., P. Makar, and D. Plummer (2012), Projections of mid-century summer air-quality

for North America: effects of changes in climate and precursor emissions, Atmospheric

Chemistry and Physics, 12(12), 5367-5390, doi:10.5194/acp-12-5367-2012. Kim, S., A. Heckel, G. Frost, A. Richter, J. Gleason, J. Burrows, S. McKeen, E. Hsie, C.

Granier, and M. Trainer (2009), NO2 columns in the western United States observed from space and simulated by a regional chemistry model and their implications for NOx emissions, Journal of Geophysical Research-Atmospheres, 114, doi:10.1029/2008JD011343.

Kim, Y., Fu, J. S., and Miller, T. L. (2010), Improving ozone modeling in complex terrain at a fine grid resolution: Part I - examination of analysis nudging and all PBL schemes associated with LSMs in meteorological model, Atmos. Environ., 44, 523–532.

Lam, Y.F., J. S. Fu, S. Wu, and L. J. Mickley (2011), Impacts of future climate change and effects of biogenic emissions on surface ozone and particulate matter concentrations in the United States, Atmos. Chem. Phys., 11, 4789–4806, doi:10.5194/acp-11-4789-2011.

Lamarque, J.-F., D. E. Kinnison, P.G. Hess and F. Vitt (2008) Simulated lower stratospheric trends between 1970 and 2005: identifying the role of climate and composition changes, J.

Geophys. Res., 113, D12301, doi:10.1029/2007JD009277. Lamarque, J.-F. and S. Solomon (2010) Impact of changes in climate and halocarbons on

recent lower stratosphere ozone and temperature trends. J. Clim., 23, 2599-2611, doi: 10.1175/2010JCLI3179.1.

Lamarque, J.-F., T. C. Bond, V. Eyring, C. Granier, A. Heil, Z. Klimont, D. Lee, C. Liousse, A. Mieville, B. Owen, M. G. Schultz, D. Shindell, S. J. Smith, E. Stehfest, J. Van Aardenne, O. R. Cooper, M. Kainuma, N. Mahowald, J. R. McConnell, V. Naik, K. Riahi and D. P. van Vuuren (2010) Historical (1850-2000) gridded anthropogenic and biomass burning emissions of reactive gases and aerosols: methodology and application. Atmos.

Chem. Phys., 10, doi:10.5194/acp-10-7017-2010, 7017-7039.


Lamarque, J.-F., G. P. Kyle, M. Meinshausen. K. Riahi, S. J. Smith, D. P. van Vuuren, A. Conley, F. Vitt (2011), Global and regional evolution of short-lived radiatively-active gases and aerosols in the Representative Concentration Pathways, Climatic Change, doi:10.1007/s10584-011-0155-0.

Lamarque, J.-F, L. K. Emmons, P. G. Hess, D. E. Kinnison, S. Tilmes, F. Vitt,C. L. Heald, E. A. Holland, P. H. Lauritzen, J. Neu, J. J. Orlando, P. Rasch, G. Tyndall (2012), CAM-chem: description and evaluation of interactive atmospheric chemistry in CESM, Geosci.

Mod. Dev., 5, 369-411, doi:10.5194/gmd-5-369-2012. Leung, L., and W. Gustafson (2005), Potential regional climate change and implications to

US air quality, Geophysical Research Letters, 32(16), doi:10.1029/2005GL022911. Liang, J., and M. Jacobson (2000), Effects of subgrid segregation on ozone production

efficiency in a chemical model, Atmospheric Environment, 34(18), 2975-2982, doi:10.1016/S1352-2310(99)00520-8.

Liao, H., W. Chen, and J. Seinfeld (2006), Role of climate change in global predictions of future tropospheric ozone and aerosols, Journal of Geophysical Research-Atmospheres, 111(D12), doi:10.1029/2005JD006852.

Liu, X., R.C. Easter, S.J. Ghan, R. Zaveri, P. Rasch, X. Shi, J.F. Lamarque, A. Gettelman, H. Morrison, F. Vitt, A. Conley, S. Park, R. Neale, C. Hannay, A.M.L. Ekman, P. Hess, N. Mahowald, W. Collins, M.J. Iacono, C.S. Bretherton, M.G. Flanner, and D. Mitchell, (2012), Toward a minimal representation of aerosols in climate models: description and evaluation in the Community Atmosphere Model CAM5, Geoscientific Model

Development, 5, 709-739, doi:10.5194/Gmd-5-709-2012. Ma, P.-L., Rasch, P. J., Fast, J. D., Easter, R. C., Gustafson Jr., W. I., Liu, X., Ghan, S. J., and

Singh, B. (2013), Assessing the CAM5 physics suite in the WRF-Chem model: implementation, evaluation, and resolution sensitivity, Geosci. Model Dev. Discuss., 6, 6157-6218, doi:10.5194/gmdd-6-6157-2013.

Mayer, M., C. Wang, M. Webster, R. G. Prinn (2000), Linking local air pollution to global chemistry and climate, J. Geophys. Res., 105(D18), 22,869–22,896.

Meehl, G., C. Covey, T. Delworth, M. Latif, B. McAvaney, J. Mitchell, R. Stouffer, and K. Taylor (2007a), The WCRP CMIP3 multimodel dataset - A new era in climate change research, Bulletin of the American Meteorological Society, 88(9), 1383-+, doi:10.1175/BAMS-88-9-1383.

Meehl, G.A., T.F. Stocker, W.D. Collins, P. Friedlingstein, A.T. Gaye, J.M. Gregory, A. Kitoh, R. Knutti, J.M. Murphy, A. Noda, S.C.B. Raper, I.G. Watterson, A.J. Weaver and Z.-C. Zhao, 2007b: Global Climate Projections. In: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.

Meleux, F., F. Solomon, F. Giorgi, (2007), Increase in summer European ozone amounts due to climate change, Atmos. Environ. 41, 7577-7587.

Mickley, L., D. Jacob, B. Field, and D. Rind (2004), Effects of future climate change on regional air pollution episodes in the United States, Geophysical Research Letters, 31(24), doi:10.1029/2004GL021216.

Morrison, H., and A. Gettelman, (2008) A new two-moment bulk stratiform cloud microphysics scheme in the community atmosphere model, version 3 (CAM3). Part I: Description and numerical tests, Journal of Climate, 21, 3642-3659, doi:10.1175/2008jcli2105.1.

Murazaki, K., and P. Hess (2006), How does climate change contribute to surface ozone change over the United States?, Journal of Geophysical Research-Atmospheres, 111(D5), doi:10.1029/2005JD005873.


Nolte, C., A. Gilliland, C. Hogrefe, and L. Mickley (2008), Linking global to regional models to assess future climate impacts on surface ozone levels in the United States, Journal of

Geophysical Research-Atmospheres, 113(D14), doi:10.1029/2007JD008497. Park, S., and C.S. Bretherton, (2009), The University of Washington shallow convection and

moist turbulence schemes and their impact on climate simulations with the Community Atmosphere Model, Journal of Climate, 22, 3449-3469, doi:10.1175/2008jcli2557.1.

Prather, M., et al. (2003), Fresh air in the 21st century?, Geophysical Research Letters, 30(2), doi:10.1029/2002GL016285.

Racherla, P. N., and P. J. Adams (2006), Sensitivity of global tropospheric ozone and fine particulate matter concentrations to climate change, J. Geophys. Res., 111, D24103, doi:10.1029/2005JD006939.

Rasmussen, D., A. Fiore, V. Naik, L. Horowitz, S. McGinnis, and M. Schultz (2012), Surface ozone-temperature relationships in the eastern US: A monthly climatology for evaluating chemistry-climate models, Atmospheric Environment, 47, 142-153, doi:10.1016/j.atmosenv.2011.11.021.

Sanderson, M., C. Jones, W. Collins, C. Johnson, and R. Derwent (2003), Effect of climate change on isoprene emissions and surface ozone levels, Geophysical Research Letters, 30(18), doi:10.1029/2003GL017642.

Sillman, S., J. Logan, and S. Wofsy (1990), A regional scale-model for ozone in the United-States with subgrid representation of urban and power plant plumes, Journal of Geophysical

Research-Atmospheres, 95(D5), 5731-5748, doi:10.1029/JD095iD05p05731. Sillman, S., and F. Samson (1995), Impact of temperature on oxidant photochemistry in

urban, polluted rural and remote environments, Journal of Geophysical Research-

Atmospheres, 100(D6), 11497-11508, doi:10.1029/94JD02146. Skamarock, W., and J. Klemp (2008), A time-split nonhydrostatic atmospheric model for

weather research and forecasting applications, Journal of Computational Physics, 227(7), 3465-3485, doi:10.1016/j.jcp.2007.01.037.

Steiner, A., S. Tonse, R. Cohen, A. Goldstein, and R. Harley (2006), Influence of future climate and emissions on regional air quality in California, Journal of Geophysical

Research-Atmospheres, 111(D18), doi:10.1029/2005JD006935. Stevenson, D. S., F. J. Dentener, M. G. Schultz, et al. (2006), Multimodel ensemble

simulations of present-day and near future tropospheric ozone, J. Geophys. Res., 111, D08301, doi:10.1029/2005JD006338.

Tagaris, E., K. Manomaiphiboon, K. Liao, L. Leung, J. Woo, S. He, P. Amar, and A. Russell (2007), Impacts of global climate change and emissions on regional ozone and fine particulate matter concentrations over the United States, Journal of Geophysical Research-

Atmospheres, 112(D14), doi:10.1029/2006JD008262. Tao, Z., A. Williams, H.-C. Huang, M. Caughey, and X.-Z. Liang (2007), Sensitivity of U.S.

surface ozone to future emissions and climate changes, Geophys. Res. Lett., 34, L08811, doi:10.1029/2007GL029455.

Tie, X.X., S. Madronich, S. Walters, et al. (2005), Assessment of the global impact of aerosols on tropospheric oxidants, J. Geophys. Res., 110, doi:10.1029/2004JD005359.

van Vuuren DP, JA Edmonds, M Kainuma, K Riahi, AM Thomson, K Hibbard, GC Hurtt, T Kram, V Krey, J-F Lamarque, T Masui, M Meinshausen, N Nakicenovic, SJ Smith, and S Rose, (2011) The representative concentration pathways: an overview Climatic Change, 109: 5-31. DOI: 10.1007/s10584-011-0148-z

Weaver, C., et al. (2009), A preliminary synthesis of modeled climate change impacts on US regional ozone concentrations, Bulletin of the American Meteorological Society, 90(12), 1843-1863, doi:10.1175/2009BAMS2568.1.

West, J., and A. Fiore (2005), Management of tropospheric ozone by reducing methane emissions, Environmental Science & Technology, 39(13), 4685-4691, doi:10.1021/es048629f.


West, J., A. Fiore, L. Horowitz, and D. Mauzerall (2006), Global health benefits of mitigating ozone pollution with methane emission controls, Proceedings of the National Academy of

Sciences of the United States of America, 103(11), 3988-3993, doi:10.1073/pnas.0600201103.

Wiedinmyer, C., S. Akagi, R. Yokelson, L. Emmons, J. Al-Saadi, J. Orlando, and A. Soja (2011), The Fire INventory from NCAR (FINN): a high resolution global model to estimate the emissions from open burning, Geoscientific Model Development, 4(3), 625-641, doi:10.5194/gmd-4-625-2011.

Wu, S., L. Mickley, D. Jacob, D. Rind, and D. Streets (2008a), Effects of 2000-2050 changes in climate and emissions on global tropospheric ozone and the policy-relevant background surface ozone in the United States, Journal of Geophysical Research-Atmospheres, 113(D18), doi:10.1029/2007JD008917|10.1029/2007JD009639.

Wu, S., L. Mickley, E. Leibensperger, D. Jacob, D. Rind, and D. Streets (2008b), Effects of 2000-2050 global change on ozone air quality in the United States, Journal of Geophysical

Research-Atmospheres, 113(D6), doi:10.1029/2007JD008917. Wu, S., L. J. Mickley, J. O. Kaplan, D.J. Jacob (2012), Impacts of changes in land use and

land cover on atmospheric chemistry and air quality over the 21st century, Atmos. Chem. Phys., 12, 1597–1609 www.atmos-chem-phys.net/12/1597/2012/doi:10.5194/acp-12-1597-2012.

Young, P., et al. (2013), Pre-industrial to end 21st century projections of tropospheric ozone from the Atmospheric Chemistry and Climate Model Intercomparison Project (ACCMIP), Atmospheric Chemistry and Physics, 13(4), 2063-2090, doi:10.5194/acp-13-2063-2013.

Yu, S., R. Mathur, G. Sarwar, D. Kang, D. Tong, G. Pouliot, and J. Pleim (2010), Eta-CMAQ air quality forecasts for O-3 and related species using three different photochemical mechanisms (CB4, CB05, SAPRC-99): comparisons with measurements during the 2004 ICARTT study, Atmospheric Chemistry and Physics, 10(6), 3001-3025.

Zeng, G., and J. Pyle (2003), Changes in tropospheric ozone between 2000 and 2100 modeled in a chemistry-climate model, Geophysical Research Letters, 30(7), doi:10.1029/2002GL016708.

Zhang, Y., X. Hu, L. Leung, and W. Gustafson (2008), Impacts of regional climate change on biogenic emissions and air quality, Journal of Geophysical Research-Atmospheres, 113(D18), doi:10.1029/2008JD009965.

Zhang, Y., Hu, X.-M., Leung, L. R., and Gustafson Jr., W. I. (2008), Impacts of regional climate change on biogenic emissions and air quality, J. Geophys. Res.-Atoms., 113, D18310, doi:10.1029/2008JD009965.

Zhang, Y., S. Olsen, and M. Dubey (2010), WRF/Chem simulated springtime impact of rising Asian emissions on air quality over the US, Atmospheric Environment, 44(24), 2799-2812, doi:10.1016/j.atmosenv.2010.05.003.


Table 1: List of NRCM-Chem simulations

SIMPRES

SIMFUT

SIMFUT_CLIM

Years 1996-2008 2046-2058 2046-2058

Climate 1996-2008 2046-2058 2046-2058

Anth. Emis. Present (2000) Future (RCP 8.5 2050) Present (2000)

Chem. IC/BC 2000 2050 2050

Methane 1.8 ppm 2.8 ppm 2.8 ppm


Table 2: Statistics for 2 m temperature (T2), water vapor mixing ratio (Q2) and total

rainfall for NRCM and NRCM-Chem over the contiguous U.S. for the present time

period (top) and future time period (bottom). Statistics are derived for the 10 years of

data where both NRCM and NRCM-Chem simulations are available. For T2 and Q2

statistics are calculated over all 6-hour values during JJA, while for rainfall the area

mean total over the 10 years and over JJA is used

PRESENT Mean T2 (C) Mean Q2 (g/kg) Total Rainfall (mm)

NRCM NRCMChem NRCM NRCMChem NRCM NRCMChem Mean 21.3 21.1 9.6 8.5 226 168 Std. Dev. 7.6 7.2 4.1 3.6 112 86 5th Perc. 9.1 8.9 3.9 3.4 46 38

95th Perc 34.2 33.0 17.4 15.0 402 306

Min -13.1 -17.3 0.0 0.0 3 4 Max 51.1 50.2 30.5 36.4 792 547

FUTURE Mean T2 (C) Mean Q2 (g/kg) Total Rainfall (mm)

NRCM NRCMChem NRCM NRCMChem NRCM NRCMChem Mean 22.9 22.8 10.7 9.6 236 176 Std. Dev. 7.7 7.3 4.6 3.8 121 96 5th Perc. 10.4 10.7 4.1 3.7 40 30 95th Perc 36.0 35.0 19.3 16.0 435 339 Min -6.1 -9.8 0.0 0.0 1 2 Max 53.0 52.5 31.9 39.0 813 610


Table 3: Anthropogenic and fire emissions for JJA over the domain. (Molecular

weight for NMVOC=72g/mol)

Anthropogenic Emissions Fires PRES FUT (FUT-PRES)/PRES CO (in Gg C) 14259 2804 -80% 1262

SO2 (in Gg S) 3415 675 -80% 14

NO (in Gg N) 2350 1020 -57% 46

NMVOC 5910 1845 -69% 645


Table 4: Statistics (Mean, Median, Standard Deviation, 5th and 95th percentiles) over

the contiguous U.S. for present time JJA average daytime surface temperature (TS

Mean), daily mean specific humidity (Q2), total daytime shortwave solar radiation

(SWDOWN) and daily maximum boundary layer height (PBLH) and the difference

(SIMFUT-SIMPRES). Results are for the entire contiguous U.S. without filtering for

statistical significance.

Mean Median Std. 5th Perc. 95th Perc.

TS Mean SIMPRES 295.1 295.9 5.6 286.1 301.8 [K] SIM

FUT-SIM

PRES 1.7 1.7 0.6 0.6 2.7

Q2 Mean SIMPRES 8.4 7.9 2.3 5.0 12.3 [g/kg] SIM

FUT-

SIMPRES

0.8 0.8 0.3 0.3 1.3

SWDOWN SIMPRES 459 446 63 373 581 [W/m2] SIM

FUT-SIM

PRES 6 8 10 -13 19

PBLH SIMPRES 820 800 282 277 1299 [m] SIM

FUT-

SIMPRES

-2 4 30 -46 51


Figure 1: Anthropogenic NOx emissions for SIMPRES (left) and their difference

between SIMFUT and SIMPRES (middle), and Fire emissions (right) over the NRCM-

Chem domain.

(totals over entire domain for JJA: 2.35TgN, 1.02TgN, 46Gg N)


Figure 2: Surface ozone at EPA monitoring sites from observations (1996-2008) and

SIMPRES. (a): Observed (black) and Modeled (red) mean (filled circles), median

(triangles), standard deviation (thick bars) and 10th and 90th percentile (lines) of

average hourly surface ozone across all sites. Blue squares show the spatial

correlation coefficient between modeled and observed hourly mean values (b):

Observed and modeled mean surface ozone mixing ratios (ppb) for 12-18 LT and 0-4

LT. (c) Observed (black) and modeled (red) cumulative distribution (solid lines) and

frequency distribution (dotted lines) of surface ozone for 12-18 LT (left-hand side)

and 0-4 LT (right-hand side). The number of data points is listed in the graphs.


Figure 3: Maps of daily maximum 8-hour surface ozone for JJA for (left; from top to

bottom) SIMPRES, SIMFUT and SIMFUT_CLIM and the difference (right; from top to

bottom) between SIMFUT and SIMPRES, SIMFUT_CLIM and SIMPRES and SIMFUT and

SIMFUT_CLIM. Grey shaded areas indicate regions with less than 95% significance.


Figure 4: (a) Frequency distribution of daily maximum 8-hour surface ozone (ppb) over the contiguous U.S. for JJA of all years of SIMPRES (black), SIMFUT (blue) and SIMFUT_CLIM (red). Mean, median, standard deviation and 5th and 95th percentiles for each simulation type are listed in the graph.


Figure 5: Average number of exceedance days for a 65 ppb standard (left) and 75 ppb

standard (right) during JJA for SIMPRES (top row), and the change in the average

number of exceedance days for SIMFUT (middle row) and SIMFUT_CLIM (bottom row).


Figure 6: Average for JJA present time absolute values (left) and difference between

SIMFUT and SIMPRES (right) of daytime surface temperature (a), daily mean specific

humidity (b), daytime downward shortwave radiation (c), daily total biogenic isoprene

emissions (d) and daily maximum boundary layer height (e). Areas with less than

95% significance are masked in the difference plots.


Figure 7: Top: Number of 1-day stagnation events during JJA for SIMPRES and the

difference between SIMFUT and SIMPRES. Bottom: Average number of unvented hours

per day during JJA for SIMPRES and the difference between SIMFUT and SIMPRES.


Figure 8: Top: Spatial distribution of the average daytime surface ozone mixing ratios

(16-22 UTC) from CAM-Chem and NRCM-Chem for present time (left) and the

difference between future and present time ozone (right). Bottom: Frequency

distributions of surface daytime ozone for CAM-Chem (blue) and NRCM-Chem

(black) over the contiguous U.S. (land; 28-50N) for present time (left), future

(middle) and the difference between future and present (right). Note the logarithmic

scale.


Figure 9: as Figure 8 but for nighttime (6-10 UTC).


Figure 10: as Figure 8, but for the Four-Corner States and local afternoon (21-22

UTC). The large cities (Den-Denver, SLC – Salt Lake City, Phx – Phoenix, Alb -

Albuquerque) are indicated in the maps.


Figure 11: as Figure 10, but for local nighttime (8-9 UTC).

Documents

Projections of Future Summertime Ozone over the U.S