Prepared By:

Abstract. The effects of future climate and emissions-related perturbations on ozone air quality in Southern California are considered, with an assumed increase to 2× pre-industrial levels for global background levels of carbon dioxide. Effects of emission and climate-related forcings on air quality are superimposed on a summer 2005 high-ozone time period. Perturbations considered here include (a) effect of increased temperature on atmospheric reaction rates, (b) effect of increased temperature on biogenic emissions, (c) effect of increased water vapor concentrations, (d) effect of increased pollutant levels at the inflow (western) boundary, and (e) effect of population growth and technology change on emissions within Southern California. Various combinations of the above perturbations are also considered. The climate-related perturbations (a–c) led to combined peak 1-h ozone increases of up to 11 ppb. The effect on ozone was greatly reduced when the temperature increase was applied mostly during nighttime hours rather than uniformly throughout the day. Increased pollutant levels at the inflow boundary also led to ozone increases up to 5 ppb. These climate and inflow-related changes offset some of the anticipated benefits of emission controls within the air basin.


Introduction
Ozone air quality varies depending on meteorological conditions. A positive correlation between ozone levels and temperature is a well-known aspect of this relationship (Aw and Kleeman, 2003;Clark and Karl, 1982;Lin et al., 2001;Sillman and Samson, Introduction Conclusions References Tables  Figures   Back  Close Full Screen / Esc

Printer-friendly Version
Interactive Discussion ozone decreases in remote areas as destruction of ozone by water vapor increases with higher temperatures. The global models do not provide a consistent picture of the response of urban ozone levels to climate change.
Regional air quality models have been used to assess local air quality effects while incorporating global-scale changes. Racherla and Adams (2008) reported longer fu- 5 ture ozone seasons and increases in summertime ozone levels in the eastern US using a "unified" model that allowed them to incorporate air quality impacts from climate change that occur outside their US study region. Jacobson (2008) used a global model with a nested regional US grid and found adverse effects on public health relating to climate change-induced air quality changes. 10 High-resolution air quality models are needed to study local effects on air quality, such as complex terrain, spatially resolved emissions, and fine-scale patterns of expected population growth. Studies such as Aw and Kleeman (2003), Steiner et al. (2006), and Kleeman (2008) refine the scale of interest to focus on air quality in urban areas. These studies incorporate predictions of future temperature changes and 15 various feedbacks that may result. Effects on ozone varied strongly by location across each air basin, highlighting the importance of using a fine-scale grid for this type of analysis.
Other studies have examined the effects on air quality of changing the frequency of stagnation events. Mickley et al. (2004) report that changes in weather patterns 20 (specifically the frequency of low-pressure systems passing through) lead to a lengthening of stagnation events across the eastern and mid-western US, creating longer and more severe ozone air pollution episodes. Similarly, Leung and Gustafson (2005) find evidence of increasing stagnation in southern California during the fall, also leading to adverse effects on ozone air quality. 25 Duffy et al. (2006) discuss increases in observed surface temperatures for California over the last 50 years, pointing out that increases in temperature have been higher during the nighttime than daytime. They argue that global climate models do not represent accurately the seasonal or diurnal changes in the observed temperature record in Cal-  Duffy et al. (2006) show that unmodeled forcings due to changes in land-use and irrigation affect trends in daily maximum temperatures. Bonfils and Lobell (2007) and Lobell and Bonfils (2008) show that increases in irrigation have had a cooling effect on daytime temperatures during summer. This daytime cooling effect may have mitigated some effects of climate change on ozone air quality.

5
The objective of this research is to predict potential effects of future changes in climate, population, and emissions on ozone air quality in southern California. Outcomes at different locations throughout the Los Angeles area are evaluated to assess interactions among the above factors.

10
The effects on ozone air quality of changes in five different factors (temperature, humidity, biogenic emissions, inflow boundary conditions, anthropogenic emissions) are evaluated in southern California. Each of these factors is considered individually. In addition, the combined air quality effects of changes in temperature, biogenic emissions, and humidity represent an aggregate climate-related effect on air quality. The 15 combined effects on air quality of future changes in anthropogenic emissions and inflow boundary conditions reflect effects of population growth and technology change, occurring locally in southern California and globally. Each of these effects on air quality is evaluated relative to a base case high-ozone episode from summer 2005.
Air pollution formation and transport is modeled for a base case period of 14-19 July Interactive Discussion tical dimension is divided into 25 layers extending to ∼15 km above sea level, with a telescoping vertical grid starting at 36 m layer thickness at ground level and increasing to 6.5 km near the tropopause. The Community Multiscale Air Quality model (CMAQ; Byun and Schere, 2006) version 4.6 is used to predict base case and future air quality. The SAPRC99 chemical 5 mechanism (Carter, 2000), containing 77 model species (including lumped VOC categories) and 216 reactions, is used to describe relevant atmospheric chemistry. Boundary conditions similar to Steiner et al. (2006) were used here except that the NO and NO 2 inflow (western) boundary conditions were reduced from 1 ppb each to 0.01 and 0.03 ppb, respectively, based on Nowak et al. (2004). Meteorological fields were developed by the South Coast Air Quality Management District (SCAQMD, 2007) using the National Center for Atmospheric Research Mesoscale Meteorological model (MM5) version 3.6.1. Three two-way nested domains were used with spatial resolution of 45, 15, and 5 km. Only the innermost 5 km meteorological fields were used in the present study to drive air quality model calculations.

15
The base case emission inventory used here was provided by the California Air Resources Board (Jackson, 2007). Separate hourly and day-specific gridded estimates of mobile, point, and area source emissions were combined with biogenic emission estimates developed using the BEIGIS model (Scott and Benjamin, 2003). Table 1 shows domain-wide emission totals for NO x , non-methane organic compounds (NMOC), and 20 CO, for each category of emissions. Table 1 highlights the importance of both onroad mobile emissions as well as off-road/area emission as important sources. A high degree of emission control has been achieved on large point sources such as power plants in southern California, so these sources are of relatively minor importance here.
Perturbations to the base case air quality model were developed and applied using 25 results from a global and regional climate modeling study by Snyder et al. (2002). A doubling of preindustrial global background CO 2 levels from 280 to 560 ppm was the basis for future climate calculations. The domain used in the regional climate model was centered on California with a horizontal resolution of 40 km. This fine resolution is Interactive Discussion needed to capture the wide variety of elevations, land cover types, and microclimates that are found within California. The regional climate model was forced using globalscale predictions from the NCAR parallel climate model for the 2×CO 2 scenario. See Snyder et al. (2002) for further details. Predicted temperature increases in southern California for July range from 1.6 to 3.5 • C, with larger increases predicted to occur 5 further inland, as shown in the first panel of Fig. 2.
Perturbations to the base case were considered separately to isolate the effects of individual variables. Temperature changes affect chemical reaction rates in the model, but these changes were not linked to other variables such as wind speed or planetary boundary layer height, which remained the same as in the base case. The effect of increased temperatures on biogenic emissions of volatile organic compounds was modeled as a separate effect. Biogenic emissions of isoprene and methyl-butenol are sensitive to temperature and light, whereas terpenes are sensitive to temperature but not light. For the future temperature scenario, emissions of isoprene and terpenes were scaled using algorithms described by Guenther et al. (1993). Methyl-butenol emissions 15 were adjusted following Harley et al. (1998).
Assuming relative humidity remains constant in the future climate scenario, absolute humidity was adjusted given the new (higher) temperatures. This calculation adjusts the ratio of (g water)/(kg air) to maintain the same relative humidity under future temperature conditions. Note that in the "humidity only" scenario relative and absolute 20 humidity both increase, while temperature is held constant. For the combined climate case, relative humidity is unchanged between future and base case scenarios, as both absolute humidity and temperature increase.
Future anthropogenic emissions were estimated from current emissions, scaled to account for population growth and technology change. Population growth in California 25 through 2050 has been forecast at the county level (DOF, 2007). We assumed that improved technologies and increased regulation of air pollution sources will reduce VOC, CO, and NO x emission factors by 80% below present-day (circa 2000, already controlled) levels. Following Steiner et al. (2006), growth in the freight-transport sector is ACPD 9,2009 Climate change and air pollution in southern California Interactive Discussion specified to be twice that of other sectors and thus diesel NO x emissions are predicted to increase in some areas (due to rapid growth), whereas VOC and CO emissions generally decrease. This future emission scenario represents business as usual, and does not reflect the effects of efforts currently being discussed to reduce greenhouse gas emissions to 80% below 1990 levels by 2050. 5 Future inflow boundary conditions were adjusted following Steiner et al. (2006), using concentrations of CO, ozone, and methane for the A1B scenario in 2000 and 2050 (IPCC, 2001). In order to highlight the potential effects from increasing emissions in Asia, ozone concentrations were increased by roughly double the amount predicted in the A1B scenario, a scenario that predicts economic growth with market based ap-10 proaches to limiting greenhouse gas emissions and assumes no single energy generation technology is dominant. On the western boundary CO concentration was increased by ∼30%, methane concentration was increased by ∼40%, and ozone was increased by ∼30%. Printer-friendly Version

Results and discussion
Interactive Discussion due to their proximity to northern and southern boundaries of the modeling domain.

Effects of climate change
A series of model runs was conducted to determine the sensitivity of air quality in southern California to future changes in emissions and climate. Figures 2, 3, and 4 show differences in input data and resulting changes in ozone air quality between 5 various modeled perturbations and the summer 2005 base case. The spatial patterns of ozone changes are shown at 03:00 p.m., which is a high-ozone time of day. Time series plots showing ozone changes at all times of day are presented in supplemental information and discussed later.
The effect of increased temperatures on chemical reaction rates is to increase peak 10 ozone levels across the domain. Figure 2 shows the change in temperature and corresponding peak ozone response. Inland areas that experience larger future temperature increases and are close enough to Los Angeles to be strongly influenced by its emissions showed the strongest ozone response to temperature changes. Figure 3 shows changes in biogenic emissions and the corresponding peak ozone 15 response. Similar to the results shown in Fig. 2, Fig. 3 shows little change in ozone for the coastal areas of Orange and Los Angeles Counties and larger ozone increases further inland. The effects of biogenic emission increases are mitigated to some extent by the spatial distribution of these emissions, which occur mostly in the surrounding mountains rather than within urbanized areas. Increased biogenic emissions had the largest 20 effects on ozone in areas north and east of Los Angeles. Over much of the domain, biogenic emissions increased by 20-35% compared to the base case as a result of higher temperatures. The response of peak ozone to increased humidity is mostly positive and can be seen in Supplemental Fig. S2

Effects of emission and inflow boundary condition changes
Increased pollutant concentrations at the western inflow boundary increase peak ozone levels consistently across all sites except Palm Springs (see Fig. 4a). There are large relative increases in ozone near the coast where base case ozone levels were generally lower compared to inland locations. 5 Under the future emission scenario, emissions of NO x , VOC and CO decrease across the domain. Coastal areas see reductions of 35-50% in NO x emissions and even larger reductions in VOC. The resulting effects on ozone were mixed, due to spatial differences in ozone sensitivity to VOC versus NO x emissions. In central Los Angeles future emissions and inflow create an increase in peak ozone levels of up to 9 ppb, 10 whereas Riverside sees a similar decrease. The effects of future changes in inflow BC and local emissions taken together are shown in Fig. 4b (combined anthropogenic effects scenario). Figure 4c shows the ozone air quality impact of a combined climate forcing case, in-15 corporating changes to temperatures, absolute humidity, and biogenic emissions. Increases in peak ozone are seen across the domain, ranging from 3-15 ppb, with the largest increases occurring inland near Riverside. The combined anthropogenic effects and climate change scenario shows increases in peak ozone at most locations with decreases seen only far inland (Fig. 4d). 20 Table 2 summarizes changes in ozone between each future scenario and the base case. Combined scenarios 6-8 are not linear combinations of individual scenarios 1-5, but instead represent results of additional model runs incorporating various combinations of the individual perturbations as noted in Table 2.

Combined effects
The locations shown in Table 2 were chosen to span future air quality outcomes in 25 populated areas near Los Angeles. Figure 1 shows a map of the study area including the locations of these 5 sites. Anaheim is located in Orange County, south of Los Overall there are large changes in peak ozone due to changes in both climate and emissions. Figure 4d combines all of the model perturbations mentioned above. Ozone increases across the greater Los Angeles area, and decreases east of Riverside. The changes in peak ozone range over ±17.5 ppb. More detailed forecasts of future emis-10 sions are needed, and air quality management plans may need to be adjusted to account for ozone increases that may occur due to climate change.

Temporal patterns of ozone change
Riverside lies near the interface between positive and negative ozone outcomes under the combined scenario of future climate, emissions, and inflow boundary conditions.

15
The magnitude and even the sign of the change in ozone concentrations at Riverside are time-dependent. The ozone response differs both by time of day, and for weekdays versus the weekend. Supplemental Fig. S3 http://www.atmos-chem-phys-discuss.net/ 9/1561/2009/acpd-9-1561-2009-supplement.pdf shows predicted difference between base case and future scenario ozone concentrations by time of day at five locations. 20 Future ozone is closest to base case predictions around noon, although the sign of the change varies from day to day at Riverside. At other sites, such as Pomona and central Los Angeles, future ozone concentrations are consistently higher than in the base case. This is due to a combination of air quality penalties due to climate change and higher inflow boundary conditions, and a local effect of lower NO x emissions on 25 ozone in upwind areas, where the system is NO x -saturated.

Conclusions
References Tables  Figures   Back  Close Full Screen / Esc

Printer-friendly Version
Interactive Discussion 3.6 Future temperature change Temperature perturbations applied here so far do not consider possible time-of-day dependence. Historically (i.e., from 1950 to present day) however, observed temperature increases during summer months have been largest at nighttime hours in many areas of California, and little increase in daytime maximum temperatures has been reported 5 in the observed record . To evaluate the effect of a diurnally-varying rather than uniform temperature increase on atmospheric chemistry and biogenic VOC emissions, an alternate form of the temperature perturbation was developed. Future temperatures were recalculated as a function of time of day: Where T 0 is the original temperature in the base case scenario, t is time in hours past midnight, and ∆T is the 24-h average increase in temperature predicted by Snyder et al. (2002) for the month of the July. The amplitude of the oscillatory component (0.8) is arbitrary, but was chosen so that most but not all of the temperature increase occurs at night. Applying the past record of summer daytime cooling due to increased 15 irrigation of 0.14 to 0.25 • C per decade in California (Bonfils and Lobell, 2007) to the 2050 temperature predictions by Snyder et al. (2002) removes most but not all of the daytime temperature increases and is consistent with Eq. (1). The increase in peak ozone predicted above with a uniform temperature increase mostly disappears when the temperature increase occurs predominantly at night. Fig-20 ure 5 shows the effects on ozone of increased temperature using a diurnally-varying temperature change (Eq. 1) relative to the base case at 03:00 and 15:00 h LT on weekdays. More generally, ozone concentrations show little change at any time of day relative to the base case when temperature increases occur mostly at nighttime hours (see Supplemental Fig. S4  Similar results were found when biogenic emissions were recalculated using diurnally-varying temperature changes. Emissions of isoprene and methyl-butenol are both temperature and light-sensitive, and emissions of these compounds are zero at night. Nighttime temperature increases do increase terpene emissions, but the absolute effect of temperature increases on terpene emissions is reduced because baseline 5 temperatures are lower at night. Ozone concentrations do not increase nearly as much as when temperatures were increased uniformly throughout the day. Changes in ozone at 03:00 and 15:00 h LT due to revised estimates of biogenic VOC emission increases are shown in Supplemental Fig. S5 http://www.atmos-chem-phys-discuss.net/9/1561/ 2009/acpd-9-1561-2009-supplement.pdf. 10 The diurnal pattern of future temperature changes is important to assessments of climate change impacts on ozone air quality. Nighttime temperature increases have less effect on ozone production than similar temperature increases that occur at midday. However, the historical record of temperature increases is not necessarily a good predictor of future warming. Irrigation, which increased over the last century in Cali-15 fornia, is unlikely to continue to increase in the same manner in future years. Loss of agricultural lands to urbanization and scarcity of water may slow or reverse the expansion in irrigated lands. In fact the amount of irrigated land in California has remained relatively stable since the 1980s (Bonfils and Lobell, 2007). Thus, any past mitigating effect of increased irrigation on daytime warming is unlikely to be repeated over the 20 next 50 years.

Summary and recommendations
This study investigated ozone air quality effects of future climate change, inflow pollutant boundary conditions, and anthropogenic emissions on ozone air quality in southern California. Future temperature changes were predicted at high spatial resolution 25 in California for a scenario of 2× pre-industrial CO 2 levels. Future emissions were predicted starting from a baseline emission inventory, factoring in expected population

Conclusions
References Tables  Figures   Back  Close Full Screen / Esc

Printer-friendly Version
Interactive Discussion growth and likely advances in emission control technologies. Globally-driven climate changes led to ozone increases throughout the study domain. In this study, climate change affected ozone levels through three mechanisms: increased temperature, increased humidity, and increased biogenic VOC emissions. These effects contribute to a climate change penalty for air quality management: ad-5 ditional controls on anthropogenic emissions will be needed to offset undesired effects of climate change on ozone. Likewise, future increases in pollutants entering southern Calfornia via inflow from the Pacific Ocean will increase ozone levels, with coastal areas being especially vulnerable to this effect. The response of ozone air quality to future emission changes varied by location, with ozone increases predicted in upwind 10 areas, and ozone decreases further downwind.
Recent commitments to control greenhouse gas emissions in California may also affect criteria pollutant emissions, and this issue requires further study. As noted above, the diurnal pattern of temperature change is important in assessing climate change effects on ozone air quality. More detailed information is needed on anticipated temper-15 ature changes as a function of time of day. Other investigators are considering effects such as changes in the number of stagnation events and length of the high-ozone season (Leung and Gustafson, 2005), and effects of climate change on other important air pollutants such as fine particulate matter (Kleeman, 2008). Finally, changes in forest fires may be an important aspect of the response to changing climate (Brown et al.,20 2004), with potentially serious effects on air quality.    Fig. 2 where T increase was applied uniformly throughout the day instead of mostly at night.