Climate Effects of Emission Standards: The Case for Gasoline and

Article pubs.acs.org/est

Climate Effects of Emission Standards: The Case for Gasoline and Diesel Cars Katsumasa Tanaka,†,‡,* Terje Berntsen,†,§ Jan S. Fuglestvedt,† and Kristin Rypdal† †

CICERO (Center for International Climate and Environmental ResearchOslo) P.O. Box 1129, Blindern, N-0318 Oslo, Norway Institute for Atmospheric and Climate Science, ETH Zurich, Universitaetstrasse 16, CHN N16.1, CH-8092 Zurich, Switzerland § Department of Geosciences, University of Oslo, P.O. Box 1022 Blindern, N-0315 Oslo, Norway ‡

S Supporting Information *

ABSTRACT: Passenger transport affects climate through various mechanisms involving both long-lived and short-lived climate forcers. Because diesel cars generally emit less CO2 than gasoline cars, CO2 emission taxes for vehicle registrations and fuels enhance the consumer preference for diesel cars over gasoline cars. However, with the non-CO2 components, which have been changed and will be changed under the previous and upcoming vehicle emission standards, what does the shift from gasoline to diesel cars mean for the climate mitigation? By using a simple climate model, we demonstrate that, under the earlier emissions standards (EURO 3 and 4), a diesel car causes a larger warming up to a decade after the emissions than a similar gasoline car due to the higher emissions of black carbon and NOX (enhancing the O3 production). Beyond a decade, the warming caused by a diesel car becomes, however, weaker because of the lower CO2 emissions. As the latter emissions standards (EURO 5 and 6) are phased in, the short-term warming due to a diesel car becomes smaller primarily due to the lower black carbon emissions. Thus, although results are subject to restrictive assumptions and uncertainties, the switch from gasoline to diesel cars encouraged by CO2 taxes does not contradict with the climate mitigation focusing on long-term consequences.

1. INTRODUCTION Transportation of passengers and goods affect the climate due to the emissions of long-lived greenhouse gases (mainly CO2), short-lived aerosols, and aerosol precursors (black carbon (BC), organic carbon (OC), and SO2), and short-lived chemically active gases (NOX, CO, and hydrocarbons (HCs)), which lead to changes in greenhouse gas concentrations (mainly CH4 and O3). Emissions from road transportation constitute a significant share of the total anthropogenic greenhouse gas emissions.1−4 The mix of emissions varies strongly among the modes of transport (road, shipping, air, and rail).1,5−7 Within the category of passenger cars, while gasoline cars release more CO and HCs, diesel cars emit more particulate matter (PM) and NOX.8 The governments of Denmark, France, The Netherlands, Norway, and the U.K. have implemented registration taxes for new cars based on CO2 emissions. The European Union as a whole has set a goal to limit the average CO2 emissions from new cars below 120 gCO2/km by 2012.9 The feasibility of the target of reaching 70 gCO2/km by 2025 is currently being assessed.9 Many states in the U.S. have already adopted, or are in the process of adopting, California’s emissions standards for new vehicles.10 In Japan, the automobile CO2 emission regulations focus on energy efficiency improvement called “Top Runner Standards”, which set mandatory energy efficiency targets according the most efficient products on the market (e.g., refs 11,12). © 2012 American Chemical Society

A climate mitigation policy that focuses on CO2 emissions supports an increase in the fraction of diesel cars because passenger cars equipped with diesel engines emit on average 15% less CO2 per kilometer than vehicles with gasoline engines with a similar power output (Tables 3−11 and 3−13 of ref 8). After the U.K. had begun taxing vehicles according to CO2 emissions, the share of registered new diesel cars increased from 26% in 2002 to 38% in 2005.13 However, it is not clear whether the shift from gasoline to diesel cars, which is encouraged by CO2 taxes for vehicle registrations and fuels, would be effective as originally intended as a climate mitigation measure, given a number of non-CO2 components emitted together.14−16 Many of them have indirect climate effects through chemical reactions or by changing the planetary albedo. NOX, CO, and HC emissions lead to a production of O3 (“short-term O3 formation”) as well as a change in the levels of the main oxidant in the atmosphere, the OH radical. NOX emissions and the produced O3 enhance the production of OH, while CO and HC emissions reduce OH levels. The change in the OH concentration affects the lifetime of CH4, which has a longer-term impact on the O3 production (“primary mode”) (ref 17 and references therein; see Received: Revised: Accepted: Published: 5205

November 22, 2011 March 12, 2012 March 21, 2012 March 21, 2012 dx.doi.org/10.1021/es204190w | Environ. Sci. Technol. 2012, 46, 5205−5213

Environmental Science & Technology

Article

Table 1. Average Emissions Used in This Study Based on the EU Legislation Emission Limits for Passenger Carsa CO2 g(CO2)/km EURO 3 (2000) EURO 4 (2005) EURO 5 (2009) EURO 6 (2014) a

gasoline diesel gasoline diesel gasoline diesel gasoline diesel

173 138 167 138 161 137 120 108

BC kg/km 2.0 2.8 2.0 1.4 1.5 2.8 1.5 2.8

× × × × × × × ×

OC kg/km

10−6 10−5 10−6 10−5 10−6 10−6 10−6 10−6

2.1 9.0 2.1 4.5 1.5 9.0 1.5 9.0

× × × × × × × ×

10−6 10−6 10−6 10−6 10−6 10−7 10−6 10−7

CO kg/km 2.3 6.4 1.0 5.0 1.0 5.0 1.0 5.0

× × × × × × × ×

10−3 10−4 10−3 10−4 10−3 10−4 10−3 10−4

HC kg/km 1.8 9.8 9.0 6.8 6.1 6.8 6.1 6.6

× × × × × × × ×

10−4 10−5 10−5 10−5 10−5 10−5 10−5 10−5

NOx kg(NO2)/km 1.5 5.0 8.0 2.5 6.0 1.8 6.0 8.0

× × × × × × × ×

10−4 10−4 10−5 10−4 10−5 10−4 10−5 10−5

See text for explanations.

Supporting Information (S1) and (S2)). BC and OC absorbs and reflects sunlight, respectively (direct effect) and BC changes the albedo of snow/ice18,19 (albedo effect). BC and OC may also change cloud properties through the indirect and semidirect effects (p 558 of ref 20). Furthermore, aerosol abundances are influenced by OH through gas-aerosol interactions.21 Added to the complex interplay discussed above, the emissions of non-CO2 components have been and will be changed under the previous and new emission standards, which further influences the complex interactions described above. In Europe, a series of regulations EURO 3 (from 2000), EURO 4 (from 2005), EURO 5 (from 2009), and EURO 6 (expected from 2014) set progressive targets for NOX, CO, HC, and PM separately for gasoline and diesel passenger cars (Table 1). Note that CO2 emissions are not regulated through the same legislation. Progressive standards have been set in Japan and the US at a rate similar to that in Europe, but the standards in the rest of the world (e.g., Russia, China, India, Brazil, and Indonesia) lag behind (Table 3 of ref 22). To analyze whether a policy that encourages the penetration of diesel cars (e.g., a registration tax of new cars based on their CO2 emissions and/or a fuel tax on CO2 emissions) is consistent with a long-term climate goal to cap the global warming, we calculate the global-mean temperature change that would be caused by the use of gasoline and diesel passenger cars under the previous and upcoming emission standards in Europe (EURO 3, 4, 5, and 6). Our method is based on a simple climate and carbon cycle model combined with aggregated parametrizations of atmospheric chemistry processes (explicitly including the short-term O3 response and the primary mode perturbation). To keep our analysis simple, gasoline and diesel cars are assumed to follow exactly the emission standards and to have a similar power output. Our approach implies the following: • Standard vs fleet emissions: We do not consider factors that would be relevant to an analysis for the real fleet on the road. A fleet of vehicles in a given year consists of many model year vehicles that represent different emission standards (cf. ref 23). • Other technological options: CO2 taxes also encourage a shift to cleaner vehicle types including gasoline-electric hybrid vehicles, plug-in-hybrid vehicles, pure electric vehicles, and hydrogen-fuel vehicles, but this study focuses on the shift from standard gasoline to diesel cars (cf. refs 24 and 25). • Well-to-tank vs tank-to-wheel emissions: Our analysis deals with the tailpipe (or end-of-pipe) emissions only (i.e., we do not perform any life-cycle analysis23,26). Wellto-tank emissions do not significantly influence our

resultsCO2-equivalent well-to-tank emissions of direct greenhouse gases would constitute 14 and 16% of wellto-wheel emissions for gasoline and diesel road vehicles, respectively.4 Our study is related to the earlier discussion14,15,27−33 but differs from the previous studies14,15 mainly in the following four regards: • Unlike the previous analyses14,15 based on sustained emissions over a long time horizon (100 years), our analysis uses one-year pulse emissions as well as sustained emissions over an average car’s lifetime (15 years). • Previous studies14,15 use a single decay constant model for CO2, whereas our study adopts the IPCC Impulse Response Function (IRF) designed to represent carbon cycle processes on various time scales (p 213 of ref 20 and ref 34). • We use a substantially lower climate efficacy for BC than those in the previous14,15 and following35,36 studies. • This study considers more atmospheric oxidation chemistry processes by accounting for the primary mode involving O3 and CH4 (see Supporting Information (S1) and (S2)). Furthermore, our study focuses on the temperature change without comparing with the adverse health impact caused by the air pollution. Many of the non-CO2 species in car exhaust have harmful effects directly through emissions of SO2, NOX, and PM13 as well as indirectly through the formation of photochemical smog, including O3. Determining the weight for the harmful effects of the short-term air pollution relative to the impact of the long-term climate change is, however, a question that involves value judgments and cannot be answered by scientific research alone (e.g., ref 37).

2. METHODS We evaluate the climate effect of various species in car exhaust with respect to the warming at the time of interest. Our approach employs a simple climate model6,38 and can be described using an analogue with the Global Temperature change Potential (GTP). The GTP is an emission metric (e.g., refs 39 and 40) that gives the global-mean temperature response from a unit emission of component i (AGTPi) relative to that of CO2 (AGTPCO2) at a chosen time point.41,42 The AGTP shows how large the warming is at a point of time ahead caused by the current emissions of a component. Our use of the GTP is in contrast to the conventional approach based on the Global Warming Potential (GWP), the emission metric adopted in the Kyoto Protocol, which integrates the RF over a chosen time interval. It has been 5206

dx.doi.org/10.1021/es204190w | Environ. Sci. Technol. 2012, 46, 5205−5213


Article

argued that the integrated RF as used in the GWP is not particularly well suited, if long-term climate policies are designed to cap the warming at a certain level.37,41,43 However, as a sensitivity case, we also perform an analysis using the GWP (see Supporting Information (S3)). We conduct the numerical experiment over a period of 100 years. The climate effect of the different emissions is not constant over time because atmospheric residence times vary enormously among species and also because radiative efficiencies (RF per concentration change) vary by up to 6 orders of magnitude (BC versus CO2). A significant addition is the chemical and photochemical interactions (involving pollutants, OH, CH4, and O3) leading to the short-term O3 formation and the following primary mode perturbation, which exhibit complex temporal behaviors. Our analysis is based on one-year pulse emissions as well as sustained emissions over an average car’s lifetime (15 years) in contrast to the previous studies14,15 using sustained emissions over the entire time perspective of the analysis (100 years). Policies that influence consumer choices about the type of engine affect emissions over the car’s lifetime, but not indefinitely. In this type of analysis, it is thus more appropriate to use pulse emissions (relevant to reversible measures such as fuel tax increase) or emissions sustained only over a relevant time scale. The analysis based on one-year pulse emissions is also instrumental to provide an insight into the climatic effects of each component, and from the one-year pulse emissions different scenarios can be constructed by regarding them as a convolution of one-year pulse emissions.6 We assume that the emission levels in our calculations are equal to the standards in EU legislations (Table 1). Further details follow: • Average CO2 emissions for gasoline engines: The average CO2 emissions for EURO 3 and 4 gasoline cars are assumed equal to those for the average Norwegian fleet newly registered in the respective legislation years (173 and 167 gCO2/km, respectively). The difference in the average CO2 emissions from EURO 3 to EURO 4 is linearly extrapolated to infer the average CO2 emissions for EURO 5 (161 gCO2/km) by considering the legislation year of each standard. The average CO2 emissions for EURO 6 are set equal to 120 gCO2/km, the EU objective by 2012.9 • Average CO2 emissions for diesel engines: Diesel cars are assumed to have 20% lower average CO2 emissions for CO2 than gasoline cars in the analysis of EURO 3. 17.5%, 15%, and 10% lower average CO2 emissions are assumed for EURO 4, 5, and 6, respectively, to reflect the anticipated improvement of fuel efficiency. These assumptions are compatible with a 15% lower average CO2 emissions for the real fleet, which is derived based on the data shown in the EMEP/EEA emission inventory guidebook 2009 (Tables 3−11 and 3−13 of ref 8). The average CO2 emissions (gCO2/km) can be decomposed into the product of the fuel oxidation factor (%), the fuel carbon content (g/gallon), and the mileage (or fuel consumption) (gallon/km). Note that the emission factor (Tier 1 method of ref 8) is the product of the fuel oxidation factor (%) multiplied by the fuel carbon content (g/gallon). On the basis of the gasoline and diesel carbon contents of 2421 and 2778 g/gallon, respectively44 and a fuel oxidation factor of 100% (Tier 1

method of ref 8), 20, 17.5, 15, and 10% lower CO2 emissions for diesel cars (EURO 3, 4, 5, and 6 emission standards, respectively) correspond to 30.3, 28.1, 25.9, and 21.6% better mileages, respectively, which is compatible with the range of mileage differences (15− 30%) considered in the previous studies.14,15 • BC and OC emissions: The PM10 emission limits specified in the EU regulations are converted to BC and OC emissions based on the size and chemical speciation factors of ref 45. • Sulfur dioxide (SO2) emissions: Sulfate aerosols, which cause a cooling by scattering solar radiation and affecting cloud properties, are not included in our analysis but do not significantly influence our results. The emission level is not directly related to the technology level of the engine, but rather the sulfur content of the fuel. Although diesel fuels contained far more sulfur than gasoline in the past, the desulfurization of fuels has dramatically been improved recently (e.g., Table 3-14 of ref 8). To compute the atmospheric CO2 concentration for given CO2 emissions, we employ an IPCC IRF derived from the Bern Carbon Cycle model (Bern2.5CC) (p 213 of ref 20 and ref 34)it consists of three exponential terms to model processes operating on various time scales and a constant term to represent the virtually permanent airborne fraction.46−48 The previous studies of refs 14 and 15 in contrast, use a model with a single decay constant and do not consider the permanent airborne fraction. The studies of refs 14 and 15 estimate a range of the single decay constant against observations for the latter half of the 20th century to express the uncertainty in the CO2 adjustment time. Approaches using a single atmospheric CO2 decay term have been criticized for their limited representation of how the carbon cycle actually works (e.g., refs 49−51). Furthermore, it is well established that such a calibration approach in refs 14 and 15 is not suitable for analyses investigating a long-term decay of small CO2 emissions (such as our study). The reasons can be summarized in the following two folds: • First, the CO2 adjustment time depends on the temporal perspective chosen: the longer the time scale one looks at, the longer the CO2 adjustment time appears.51 The choice of the time horizon does hinge on the problem at hand (e.g., pp 167−173 of ref 52), however, the time window of half a century used to tune the CO2 time constant in refs 14,15 is not sufficiently long for our experiment. • Second, more importantly, the CO2 adjustment time depends on the emission trajectory used as input to the calculations:53 the larger the growth of CO2 emissions is, the shorter the CO2 adjustment time appears.54 A single decay constant derived from a system perturbed by fast growing emissions (e.g., the second half of the 20th century) reflects mainly short-term processes in the ocean mixed layer and is thus different from a CO2 adjustment time derived from a long-term pulse emission experiment driven primarily by long-term processes in the deep ocean (Point 3 of ref 55 and Table 2 of ref 56). Furthermore, the CO2 time scale depends on how it is defined (e.g., turnover time vs residence time57) and how the uncertainty in the carbon cycle is addressed (e.g., Section B.4.1 of ref 58 and refs 59 and 60). More detailed discussion on the 5207



Article

boxes: the upper box represents both the atmosphere and the mixed layer of the ocean, and the lower box represents the deep ocean. Although the variable of interest is the temperature change of the upper box, it is important to consider the thermal inertia of the deep ocean when the long-term effect of the emission of a short-lived climate forcer is compared to that of the long-lived CO2 perturbation. The climate sensitivity is assumed to be 3.3 °C in our standard setup.

carbon cycle uncertainty can be found in Supporting Information (S4). RFs following the emissions of CO2, NOX, CO, HCs, and BC and OC aerosols are calculated by applying a set of parametrizations. Note that the simple parametrizations adopted here do not capture a full suite of processes. The nitrate aerosol effect and the nitrogen fertilization effect are two cooling mechanisms that are very uncertain and not included in our study (cf. ref 21)however, both would potentially lead to a larger difference between gasoline and diesel cars. The indirect and semidirect effects are still poorly understood and not included in our analysis.35,36 • CO2 forcing: The radiative forcing of CO2 is calculated by multiplying the change in atmospheric CO 2 concentration following the emission by the radiative efficiency (p 212 of ref 20). • BC and OC forcing: The BC and OC forcing is calculated by Q(t) = Em × Q0 × e−t/τ, where Q0, τ, and Em denote the conversion factor from the emission to the radiative forcing, the lifetime, and the emission, respectively. For BC (direct effect: fossil fuel combustion), τ is 7.1 days and Q0 is 1.96 × 10−9 Wm−2kg−1, while for OC, τ is 7.6 days and Q0 is −2.88 × 10−10 Wm−2kg−1.61 For the BC forcing (albedo effect: snow and ice deposition), we assume a lifetime of one month (to account for its relatively short-term enhancement of RF), and then optimize the parameter Q0 such that the integral of Q(t) over one year is equal to the globalannual-mean RF of 0.039 Wm−2 (ref 18) when the total annual BC emissions of 5.1 Tg/year62 is applied. We explicitly consider the efficacy of BC (0.74 for the direct effect and 1.7 for the albedo effect63). The efficacies of other forcing agents are assumed to be 1.0. • Short-term O3 response: The RF caused by the shortterm O3 concentration change in response to pollutant emissions are calculated based on the parametrization formulated as a function of the CH4 concentration and the emissions of NOX, CO, and HCs, which is derived from the sensitivity analyses using chemical transport models (CTMs) (p 269 of ref 64). The conversion factor of 0.042 Wm−2DU−1 (p 402 of ref 64) is used. We use a lifetime of 0.25 years.15 • Primary mode (O3 and CH4): The calculation of the CH4 and O3 concentration changes related to the primary mode follows the methodology in Appendix B of.65 The following are some key aspects. The CH4 adjustment time with respect to the OH depletion is defined to have an inverse relationship with the relative change in the OH concentration (e.g., p 73 of ref 48). The OH concentration is computed based on the parametrization given as a function of the CH 4 concentration and the emissions of NOX, CO, and HCs (p 269 of ref 64 and p 73 of 48). A 10% increase in the CH4 concentration leads to an increase in tropospheric O3 by 0.64 DU.65 The time constant of such a change in tropospheric O3 is assumed to be the time constant of the primary mode, which is 40% longer than the CH 4 turnover time. 17 Also see Supporting Information (S1) and (S2). The RF terms described above are the input to a simple upwelling-diffusion model6,38 to compute the global-mean surface temperature change. The model consists of two

3. RESULTS 3.1. Analysis Based on One-Year Pulse Emissions. Our calculations show that the newer emission standards generally lead to a smaller net warming (Figure 1). Our results also

Figure 1. Global-mean temperature change caused by car exhaust emitted from gasoline and diesel engines specified in the previous and upcoming European vehicle emission standards (EURO 3, 4, 5, and 6). The results are based on one-year pulse emissions in year zero, which would be emitted by an idealized 1000 km drive.

demonstrate that the long-term warming from diesel engines is smaller relative to that from gasoline engines for any emission standard because only the CO2 emissions count in the long run. After 100 years, the warming from diesel cars is about 20.4% smaller than that from gasoline cars in the case of EURO 3 and 17.7%, 15.4%, and 10.2% lower in the case of EURO 4, 5, and 6, respectively, reflecting our assumptions on the average CO2 emissions (Section 2). The crossover points, beyond which diesel cars cause less warming than gasoline cars, occur closer to the emission year as the advanced emissions standards set in 5 years, 5 years, 3 years, and 2 years for EURO 3, 4, 5, and 6, respectively, because of the more stringent caps for short-lived components (O3 precursors and BC) in the newer emission standards. In the cases of EURO 5 and 6, BC emissions from diesel cars cause only a small warming. In the following, we discuss the temperature response to the emissions of each component for EURO 4 (Figure 2): • CO2 and non-CO2 agents: During the first 10 years of contributions from the non-CO2 components are significant, causing both warming and cooling. The short-term impacts (warming and cooling) are substantially larger for diesel engines due to the higher emissions of NOX (leading to short-term O3 formation) and carbonaceous particles. However, the net warming due to diesel engines becomes smaller after less than a decade as the total warming is increasingly controlled by CO2, the emissions of which are smaller from diesel engines. • BC and OC: The only significant effect of carbonaceous particles is the BC emissions from diesel engines. At the 5208



Article

Figure 2. Global-mean temperature change caused by car exhaust emitted from a single idealized car specified in EURO 4 [gasoline engines (left panel) and diesel engines (right panel)]. The results are based on one-year pulse emissions in year zero, which would be emitted by an idealized 1000 km drive. Note that the warming curve for BC (diesel) overlaps with that for the short-term O3 response (diesel).

end of the first year, the contribution from BC is as large as 43% of the total warming, but during the first decade after the emissions it decreases to 10%. In the long term (20 years or more), the contribution from BC is very small (less than 2%). The climatic effect of OC is weaker than that of BC because of the lower radiative efficiency of OC and also because of the smaller fraction of OC in PM in the case of diesel vehicles.45 • Short-term O3 response: The climate response due to the short-term increase in the O3 concentration is stronger for diesel engines than for gasoline engines. This shortterm O3 effect explains 45% of the total warming at the end of the first year. • Primary mode (O3 and CH4): The temperature change due to the primary mode for diesel engines is also stronger than that for gasoline vehiclesthe responses are, however, of the opposite sign. For diesel cars, the primary mode effect reduces the warming by up to 22% (10 years after the emissions). For gasoline cars, the lower ratio of NOX to CO + HC emissions leads to a reduction in the OH concentration and thus an increase in the CH4 concentration, resulting in a small temperature increase (less than 5% of the net warming). The primary mode has a longer lasting effect on the temperature change (i.e., an intermediate time scale) than the short-term O3 change and carbonaceous particles emissions because it affects the concentration of CH4, which has a perturbation response time of 12 years (p 212 of ref 20). 3.2. Analysis Based on 15-Year Sustained Emissions. The initial temperature response is drastically different between the early emission standards (Euro 3 and 4) and the later standards (Euro 5 and 6) (Figure 3). Under the previous standards, the magnitude of the warming caused by diesel cars exceeds by far that by gasoline cars during the early phase of operation [by up to a factor of 2.0 and 2.5 under EURO 3 and 4, respectively (Figure 4)] due to the large influence from the short-lived components O3 and BC. After about a decade (11 and 10 years under EURO 3 and 4, respectively), which is before the end of the assumed car’s lifetime (15 years), the warming caused by diesel cars becomes smaller owing to the long-term importance of CO2 and the cooling effect of NOX through CH4 reduction by the primary mode perturbation. However, under EURO 5 and 6, the emissions of short-lived components are so low that the warming due to diesel cars is surpassed by that of gasoline cars just after 4 and 2 years of

Figure 3. Global-mean temperature change caused by car exhaust emitted from gasoline and diesel engines specified in the previous and upcoming European vehicle emission standards (EURO 3, 4, 5, and 6). The results are based on sustained 15-year emissions from year zero, which would be emitted by an idealized 1000 km drive per year.

Figure 4. The ratios of the warming caused by car exhaust emitted from diesel engines to that from gasoline engines following EURO 3, 4, 5, and 6. The results are based on sustained 15-year emissions from year zero, which would be emitted by an idealized 1000 km drive per year.

operation, respectively. However, the long-term temperature response is largely unaffected by the short-term response and converges to the warming caused by the CO2 emissions in all of the emission standards (Figure 3)diesel-fueled vehicles cause a 20.5%, 17.7%, 15.4%, 10.2% smaller warming after 100 years than gasoline-fueled vehicles for the EURO 3, 4, 5, and 6 cases, respectively (Figure 4). 5209



Article

representative uncertainty range is from 2.0 to 4.5 °C per doubling CO2 concentration (pp 798−799 of ref 20). Such an uncertainty in the climate sensitivity leads to a 90% confidence interval of ±30% in the temperature response from the emissions of gasoline and diesel engines (Figures 1, 2, and 3). However, the 90% confidence interval of the ratio of the warming from diesel to that from gasoline (Figure 4) is only ±10%.

3.3. Uncertainties. Below we discuss how the uncertainties in CO2 emissions, BC forcing, short-term O3 response, and climate sensitivity influence our results. The uncertainty in the carbon cycle is explored in Supporting Information (S4). • CO2 emissions: The difference in the CO2 emissions between gasoline and diesel cars is subject to an uncertainty. We performed a sensitivity analysis by assuming 20% and 35% better mileages for diesel cars in all of the emission standards. With a 20% better mileage for diesel cars, the crossover points based on 15year sustained emissions (Figure 4), at which the warming due to diesel cars is surpassed by that due to gasoline cars, are shifted to 13, 13, 6, and 3 years for EURO 3, 4, 5, and 6, respectively. Using a 35% better mileage leads to the crossover points at 10, 8, 3, and 1 year(s) for EURO 3, 4, 5, and 6, respectively (compare the mileage difference in the reference case in Section 2). • BC forcing: The climate impact of BC emissions is still quite uncertain (p 164 of ref 20 and ref 66). The previous studies14,15 use a much higher climate efficacy to include a number of indirect effects, in accordance with more detailed BC-cloud interactions.67 A sensitivity analysis is performed by using a higher estimate of the BC forcing, as well as the efficacy, to quantify possible impacts of uncertainties in forcing and response to BC emissions. We assume a maximum specific forcing and the lifetime corresponding to a global direct RF of 1.0 Wm−2 (ref 66) and double the RF for the snow-albedo effect to be equal to the estimate (0.08 Wm−2).68 We also increase the efficacy for the direct effect from 0.74 to 1.3 (Figure 2.19 of ref 20), and for the snow/albedo effect from 1.7 to 4.5 (maximum efficacy from ref 19). These assumptions have a sizable effect on the short-term temperature response but hardly influence the long-term response. On the basis of one-year pulse emissions from diesel engines following EURO 4, the fraction of warming that the BC forcing accounts for is 77% initially, 35% after a decade, and 3% after 3 decades (compared to 43%, 10%, and 0.5% in the standard results). • Short-term O3 response: The simple linear IPCC formulas applied here do not account for the fact that the O3 production efficiency per NOX molecule oxidized decreases with increasing background NOX levels (e.g., ref 69). A large share of the actual emissions from road traffic takes place in polluted regions with high abundances of NOX. Although accounting for polluted background concentrations is beyond the scope of our analysis assuming an idealized driving condition, we compare our O3 forcing estimate with the estimate obtained from a global 3-D CTM1 and detailed radiative forcing calculations. The model calculations give an estimate of 6.2 mWm−2/Tg(N)/year, while an estimate of 8.5 mWm−2/Tg(N)/year is derived from our linear parametrization (based on the assumption that the gasoline usage is about 60% of the total fuel used for road transport70). This indicates that our estimate of the warming from O3 in the early phase, which is a factor of 2.5 stronger for diesel engines, is 35% higher than the estimate in ref 1. • Climate sensitivity: A key uncertainty in the global climate system is the climate sensitivity (e.g., ref 71). Its

4. DISCUSSION It has been debated whether policies (e.g., CO2 taxes for new car registrations and fuels) that encourage a shift from gasolinefueled cars to diesel-fueled cars would be effective as a measure to mitigate the global climate change (e.g., refs 14, 15, and 27−33). We demonstrated that the shift from gasoline to diesel engines following the previous and upcoming European emission standards is consistent with long-term climate mitigation efforts. The previous studies14,15 show more prolonged warmings due to diesel cars under various conditionsthe difference between the earlier results and our results can be explained mainly by the combination of the factors discussed above such as underlying emissions (pulse or sustained), carbon cycle model, BC efficacy, and atmospheric chemistry processes considered. While our analysis is useful to advance the scientific understanding of the climatic effects implied by the vehicle emission standards, our results cannot be directly applied to the analysis of real fleet. The actual emissions depend on a number of factors such as driving conditions (urban, suburban, and rural), vehicle specifications [engine size, engine types (e.g., gasoline direct injection), aerodynamic drag, rolling resistance, vehicle weight, particle trap, etc.], the model mix on the road, and the lifetimes of individual vehicles. It is also known that particular vehicles (e.g., those lacking maintenance) are responsible for a substantial fraction of the on-the-road emissions, termed “skewness of the emissions”.45,72,73 The on-road traffic consists of not only new cars (analyzed here) but also used carsnew types of vehicles such as gasoline− electric hybrid vehicles, plug-in-hybrid vehicles, pure electric vehicles, and hydrogen-fuel vehicles, which are not a subject of our study (cf. refs 24 and 25) but may well be cleaner options, are increasing in numbers in recent years. Furthermore, unlike our conceptual approach using one-year pulse or 15-year sustained emissions, the emissions continue in the real fleet without being cut off at a certain point in the futurehowever, an analysis of the long-term real fleet would face additional uncertainties related to the future changes in fleet and emission composition along with the technological change. When interpreting our conclusion that CO2 taxes are in line with efforts to reduce our impact on the climate, one should keep in mind that in general, considering relevant non-CO2 components in addition to CO2 is essential in designing climate-related policy instruments because non-CO2 components play an important role in characterizing the short- to middle-term climate responseit is not unlikely that other instruments in different sectors or cases lead to a variety of climatic outcomes including those incompatible with efforts to reduce our impact on the climate. Furthermore, policy makers would need additional considerations beyond what we addressed here. For example, the rebound effect74,75 may significantly influence our results. It has been argued that consumers use a part of the money saved through opting for a diesel engine for upgrading to a larger engine. The rebound 5210



Article

(5) Lauer, A.; Eyring, V.; Hendricks, J.; Jöckel, P.; Lohmann, U. Global model simulations of the impact of ocean-going ships on aerosols, clouds, and the radiation budget. Atmos. Chem. Phys. 2007, 7 (19), 5061−5079. (6) Berntsen, T.; Fuglestvedt, J. Global temperature responses to current emissions from the transport sectors. Proc. Natl. Acad. Sci. 2008, 105 (49), 19154−19159. (7) Grewe, V.; Stenke, A. AirClim: An efficient tool for climate evaluation of aircraft technology. Atmos. Chem. Phys. 2008, 8 (16), 4621−4639. (8) European Environment Agency (EEA). EMEP/EEA Air Pollutant Emission Inventory Guidebook2009. In EEA Technical Report. (9) European Commission. RoadTransport: Reducing CO2 Emissions from Light-Duty Vehicles; 2010. (10) PEW Center on Global Climate Change. State Legislation from around the Country; 2011. (11) Kimura, O.Japanese Top Runner Approach for Energy Efficiency Standards. In SERC Discussion Papers; Socio-Economic Research Center (SERC); 2009. (12) European Automobile Manufacturs’ Association. Japan Opts for Integrated Approach; 2007. (13) Mazzi, E. A.; Dowlatabadi, H. Air quality impacts of climate mitigation: UK policy and passenger vehicle choice. Environ. Sci. Technol. 2006, 41 (2), 387−392. (14) Jacobson, M. Z. Control of fossil-fuel particulate black carbon and organic matter, possibly the most effective method of slowing global warming. J. Geophys. Res. 2002, 107 (D19), 4410. (15) Jacobson, M. Z. Correction to “Control of fossil-fuel particulate black carbon and organic matter, possibly the most effective method of slowing global warming”. J. Geophys. Res. 2005, 110 (D14), D14105. (16) Boucher, O; Reddy, M. S. Climate trade-off between black carbon and carbon dioxide emissions. Energy Policy 2008, 36 (1), 193− 200. (17) Prather, M. J. Lifetimes and time scales in atmospheric chemistry. Philos. Trans. R. Soc. A: Math., Phys. Eng. Sci. 2007, 365 (1856), 1705−1726. (18) Warren, S. G.; Wiscombe, W. J. A model for the spectral albedo of snow. II: Snow containing atmospheric aerosols. J. Atmos. Sci. 1980, 37 (12), 2734−2745. (19) Flanner, M. G.; Zender, C. S.; Randerson, J. T.; Rasch, P. J. Present-day climate forcing and response from black carbon in snow. J. Geophys. Res. 2007, 112 (D11), D11202. (20) IPCC (2007) Climate Change: Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change; Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K. B., Tignor, M., Miller, H. L., Eds.; Cambridge University Press: Cambridge/New York, 2007; p 996. (21) Shindell, D. T.; et al. Improved attribution of climate forcing to emissions. Science 2009, 326 (5953), 716−718. (22) Rafaj, P., Rao, S., Klimont, Z., Kolp, P., Schöpp, W. Emissions of Air Pollutants Implied by Global Long-Term Energy Scenarios. In Interim Report; International Institute for Applied Systems Analysis (IIASA): Laxberg, Australia, 2010; p 28. (23) Peters, G. P.; Aamaas, B; Lund M, T.; Solli, C; Fuglestvedt, J. S. Alternative “Global Warming” metrics in life cycle assessment: A case study with existing transportation data. Environ. Sci. Technol. 2011, 45 (20), 8633−8641. (24) Samaras, C.; Meisterling, K. Life cycle assessment of greenhouse gas emissions from plug-in hybrid vehicles: Implications for policy. Environ. Sci. Technol. 2008, 42 (9), 3170−3176. (25) Peterson, S. B.; Whitacre, J. F.; Apt, J. Net air emissions from electric vehicles: The effect of carbon price and charging strategies. Environ. Sci. Technol. 2011, 45 (5), 1792−1797. (26) Mikhail, V. C.; Arpad, H. Environmental assessment of passenger transportation should include infrastructure and supply chains. Environ. Res. Lett. 2009, 4 (2), 024008. (27) Feichter, J.; Sausen, R.; Graßl, H.; Fiebig, M. Comment on “Control of fossil-fuel particulate black carbon and organic matter,

effect may also work through increased driven mileage (in addition to larger engines). This implies that the difference between the CO2 emissions from gasoline engines and those from diesel engines could be effectively smaller than what we obtained, reducing further the difference between the long-term climate effects expected for gasoline-fueled vehicles and that for diesel-fueled vehicles. Under the limitations discussed above, we demonstrated that CO2 taxes clearly lead to a larger short-term warming from diesel-fueled cars under the earlier emission standards (EURO 3 and 4). However, as the newer emission standards (EURO 5 and 6) are progressively set in, such a short-term warming becomes negligible. As poorly maintained vehicles may not follow the emission standards, our findings emphasize that mitigation measures that focus on CO2 (and thus may lead to a larger fraction of diesel cars) need to be supplemented by a measure to enforce proper maintenance to ensure that the vehicles are kept up to the standard. This is particularly important for diesel cars since the potential for high BC emissions with enhanced short-term warming effects is substantial.

■

ASSOCIATED CONTENT

S Supporting Information *

Details of the primary mode, NOX and CO impact on the primary mode and tropospheric O3, the Kyoto Protocol perspective, and sensitivity analysis of the CO2 pulse response. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +41 44 632 34 95; fax: +41 44 633 10 58; e-mail: [email protected] Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Marianne T. Lund, colleagues in the QUANTIFY and TEMPO projects, and three anonymous reviewers for valuable and in-depth comments and discussions. This research was funded by the European Union’s Sixth Framework Program Project QUANTIFY under Contract No. 003893 and by the Norwegian Research Council under project 195191/ S60 (TEMPOTransport and Environment−Measures and Policies). K.T. is supported by the Norwegian Research Council under Project 203807 (ClimUpClimate feedback uncertainty and its policy implications) and by the Marie Curie Intra-European Fellowship within the seventh European Community Framework Programme (Proposal No. 255568 under FP7-PEOPLE-2009-IEF).

■

REFERENCES

(1) Fuglestvedt, J; Berntsen, T; Myhre, G; Rypdal, K; Skeie, R. B. Climate forcing from the transport sectors. Proc. Natl. Acad. Sci. 2008, 105 (2), 454−458. (2) Skeie, R. B.; et al. Global temperature change from the transport sectors: Historical development and future scenarios. Atmos. Environ. 2009, 43 (39), 6260−6270. (3) Unger, N.; et al. Attribution of climate forcing to economic sectors. Proc. Natl. Acad. Sci. 2010, 107 (8), 3382−3387. (4) Uherek, E.; et al. Transport impacts on atmosphere and climate: Land transport. Atmos. Environ. 2010, 44 (37), 4772−4816. 5211



Article

possibly the most effective method of slowing global warming” by M. Z. Jacobson. J. Geophys. Res. 2003, 108 (D24), 4767. (28) Chock, D. P.; Song, Q.; Hass, H.; Schell, B.; Ackermann, I. Comment on “Control of fossil-fuel particulate black carbon and organic matter, possibly the most effective method of slowing global warming” by M. Z. Jacobson. J. Geophys. Res. 2003, 108 (D24), 4769. (29) Penner, J. E. Comment on “Control of fossil-fuel particulate black carbon and organic matter, possibly the most effective method of slowing global warming” by M. Z. Jacobson. J. Geophys. Res. 2003, 108 (D24), 4771. (30) Jacobson, M. Z. Reply to comment by J. E. Penner on “Control of fossil-fuel particulate black carbon and organic matter, possibly the most effective method of slowing global warming”. J. Geophys. Res. 2003, 108 (D24), 4772. (31) Jacobson, M. Z. Reply to comment by D. P. Chock et al. on “Control of fossil-fuel particulate black carbon and organic matter, possibly the most effective method of slowing global warming”. J. Geophys. Res. 2003, 108 (D24), 4770. (32) Jacobson, M. Z. Reply to comment by J. Feichter et al. on “Control of fossil-fuel particulate black carbon and organic matter, possibly the most effective method of slowing global warming”. J. Geophys. Res. 2003, 108 (D24), 4768. (33) Jacobson, M. Z. Correction to “Reply to comment by J. Feichter et al. on 'Control of fossil-fuel particulate black carbon and organic matter, possibly the most effective method of slowing global warming'”. J. Geophys. Res. 2003, 108 (D24), 4806. (34) Joos, F; et al. An efficient and accurate representation of complex oceanic and biospheric models of anthropogenic carbon uptake. Tellus B 1996, 48, 397−417. (35) Koch, D; Del Genio, A. D. Black carbon semi-direct effects on cloud cover: Review and synthesis. Atmos. Chem. Phys. 2010, 10 (16), 7685−7696. (36) Koch, D; et al. Soot microphysical effects on liquid clouds, a multi-model investigation. Atmos. Chem. Phys. 2011, 11 (3), 1051− 1064. (37) Berntsen, T.; Tanaka, K.; Fuglestvedt, J. Does black carbon abatement hamper CO2 abatement? Climatic Change 2010, 103 (3), 627−633. (38) Schneider, S. H.; Thompson, S. L. Atmospheric CO2 and climate: Importance of the transient response. J. Geophys. Res. 1981, 86 (C4), 3135−3147. (39) Fuglestvedt, J. S.; et al. Metrics of climate change: Assessing radiative forcing and emission indices. Climatic Change 2003, 58 (3), 267−331. (40) Tanaka, K; Peters, G. P.; Fuglestvedt, J. S. Policy Update: Multicomponent climate policy: Why do emission metrics matter? Carbon Manag. 2010, 1 (2), 191−197. (41) Shine, K. P.; Berntsen, T. K.; Fuglestvedt, J. S.; Skeie, R. B.; Stuber, N. Comparing the climate effect of emissions of short- and long-lived climate agents. Philos. Trans. R. Soc. A: Math., Phys. Eng. Sci. 2007, 365 (1856), 1903−1914. (42) Shine, K. P.; Fuglestvedt, J. S.; Hailemariam, K.; Stuber, N. Alternatives to the Global Warming Potential for comparing climate impacts of emissions of greenhouse gases. Climatic Change 2005, 68 (3), 281−302. (43) Manne, A. S.; Richels, R. G. An alternative approach to establishing trade-offs among greenhouse gases. Nature 2001, 410 (6829), 675−677. (44) Environmental Protection Agency. Emission Facts: Average Carbon Dioxide Emissions Resulting from Gasoline and Diesel fuel. Office of Transportation and Air Quality at U. S. Environmental Protection Agency; 2005. (45) Bond, T. C.; et al. A technology-based global inventory of black and organic carbon emissions from combustion. J. Geophys. Res. 2004, 109 (D14), D14203. (46) Lasaga, A. C. The kinetic treatment of geochemical cycles. Geochim. Cosmochim. Acta 1980, 44 (6), 815−828.

(47) Hooss, G.; Voss, R.; Hasselmann, K.; Maier-Reimer, E.; Joos, F. A nonlinear impulse response model of the coupled carbon cycleclimate system (NICCS). Climate Dyn. 2001, 18 (3), 189−202. (48) Tanaka, K., et al. Aggregated Carbon Cycle, Atmospheric Chemistry, and Climate Model (ACC2)Description of the Forward and Inverse Modes. In Reports on Earth System Science; Max Planck Institute for Meteorology: Hamburg, 2007; p 188. (49) Lashof, D. A.; Ahuja, D. R. Relative contributions of greenhouse gas emissions to global warming. Nature 1990, 344 (6266), 529−531. (50) Tans, P. P. The CO2 lifetime concept should be banished; an editorial comment. Climatic Change 1997, 37 (3), 487−490. (51) Archer, D; et al. Atmospheric lifetime of fossil fuel carbon dioxide. Ann. Rev. Earth Planet. Sci. 2009, 37 (1), 117−134. (52) Mackenzie, F. T. Our Changing Planet: An Introduction to Earth System Science and Global Environmental Change; Prentice Hall: Boston, 2011. (53) Tubiello, F. N.; Oppenheimer, M. Impulse-response functions and anthropogenic CO2. Geophys. Res. Lett. 1995, 22 (4), 413−416. (54) Victor, D. G. Calculating greenhouse budgets. Nature 1990, 347, 431. (55) Gaffin, S. R.; O’Neill, B. C.; Oppenheimer, M. Comment on “The lifetime of excess atmospheric carbon dioxide” by Berrien Moore III and B. H. Braswell. Global Biogeochem. Cycles 1995, 9 (1), 167− 169. (56) Maier-Reimer, E.; Hasselmann, K. Transport and storage of CO2 in the oceanAn inorganic ocean-circulation carbon cycle model. Climate Dyn. 1987, 2 (2), 63−90. (57) O’Neill, B. C.; Gaffin, S. R.; Tubiello, F. N.; Oppenheimer, M Reservoir timescales for anthropogenic CO2 in the atmosphere. Tellus B 1994, 46 (5), 378−389. (58) Tanaka, K. Inverse Estimation for the Simple Earth System Model ACC2 and Its Applications, Ph.D. Dissertation, Hamburg Universität, Hamburg, 2008. (59) Knorr, W. Is the airborne fraction of anthropogenic CO2 emissions increasing? Geophys. Res. Lett. 2009, 36 (21), L21710. (60) Friedlingstein, P.; Prentice, I. C. Carbon−climate feedbacks: A review of model and observation based estimates. Curr. Opin. Environ. Sust. 2010, 2 (4), 251−257. (61) Schulz, M.; et al. Radiative forcing by aerosols as derived from the AeroCom present-day and pre-industrial simulations. Atmos. Chem. Phys. 2006, 6 (12), 5225−5246. (62) Bond, T. C.; Sun, H. Can reducing black carbon emissions counteract global warming? Environ. Sci. Technol. 2005, 39 (16), 5921−5926. (63) Hansen, J.; Nazarenko, L. Soot climate forcing via snow and ice albedos. Proc. Natl. Acad. Sci. 2004, 101 (2), 423−428. (64) IPCC Climate Change: The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change; Houghton, J. T., Ding, Y, Griggs, D. J., Noguer, M., van der Linden, P. J., Dai, X., Maskell, K., Johnson, C. A. Eds.; Cambridge:Cambridge/New York, 2001; p 881. (65) Berntsen, T. K.; et al. Response of climate to regional emissions of ozone precursors: sensitivities and warming potentials. Tellus B 2005, 57 (4), 283−304. (66) Ramanathan, V.; Carmichael, G. Global and regional climate changes due to black carbon. Nat. Geosci. 2008, 1 (4), 221−227. (67) Jacobson, M. Z. Short-term effects of controlling fossil-fuel soot, biofuel soot and gases, and methane on climate, Arctic ice, and air pollution health. J. Geophys. Res. 2010, 115 (D14), D14209. (68) Hansen, J.; et al. Efficacy of climate forcings. J. Geophys. Res. 2005, 110 (D18), D18104. (69) Jacob, D. J. Introduction to Atmospheric Chemistry; Princeton University Press: Princeton, 1999. (70) Borken, J.; Steller, H.; Merétei, T.; Vanhove, F. Global and country inventory of road passenger and freight transportation: fuel consumption and emissions of air pollutants in year 2000. J. Trans. Res. Board 2007, 2011 (1), 127−136. 5212



Article

(71) Tanaka, K.; Raddatz, T.; O’Neill, B. C.; Reick, C. H. Insufficient forcing uncertainty underestimates the risk of high climate sensitivity. Geophys. Res. Lett. 2009, 36 (16), L16709. (72) Bishop, G. A.; Stedman, D. H. A decade of on-road emissions measurements. Environ. Sci. Technol. 2008, 42 (5), 1651−1656. (73) Ban-Weiss, G. A.; Lunden, M. M.; Kirchstetter, T. W.; Harley, R. A. Measurement of black carbon and particle number emission factors from individual heavy-duty trucks. Environ. Sci. Technol. 2009, 43 (5), 1419−1424. (74) Sorrell, S; Dimitropoulos, J; Sommerville, M Empirical estimates of the direct rebound effect: A review. Energy Policy 2009, 37 (4), 1356−1371. (75) Schipper, L; Marie-Lilliu, C; Fulton, L Diesels in Europe: analysis of characteristics, usage patterns, energy savings and CO2 emission implications. J. Trans. Econ. Policy 2002, 36 (2), 305−340.

5213


Climate Effects of Emission Standards: The Case for Gasoline and

Recommend Documents