Climate and health impacts of US emissions reductions consistent with 2 C

Transcription

1 Climate and health impacts of US emissions reductions consistent with 2 C Drew T. Shindell, Yunha Lee, Greg Faluvegi Emission Scenarios We first calculated sectoral emissions changes in 2030 that would be required for an overall national emissions trajectory consistent with 2 C. Emissions reductions over the coming decades are made all the more difficult by projected increases in population and per capita demand for services. These drivers lead to substantial increases in emissions under a business-as-usual scenario. We use the Representative Concentration Pathway (RCP) 8.5 scenario produced by the MESSAGE Integrated Assessment Model (1) as a baseline. Under that scenario, US emissions from the energy sector rise by ~2700 Tg CO 2 eq yr -1 and those from the surface transportation sector increase ~1250 Tg yr -1 (where CO 2 eq stands for CO 2 equivalent and includes all well-mixed greenhouse gases). Assuming that the smaller contribution from industry and agriculture (~2000 Tg yr -1 ) remains roughly constant, reflecting a balance between modest gains via improved industrial efficiency and agricultural management and increased demand for manufactured goods and food, this represents an increase from current (2010) US emissions of ~6700 Tg yr -1 to ~10,600 Tg yr -1. The US had pledged an 80% reduction in warming emissions by 2050 (relative to ~current), and such reductions would be consistent with 2 C. These are implemented linearly, assuming that a front-loaded trajectory such as this in which emission cuts are of greatest magnitude in absolute terms early on is among the most likely paths for drastic cuts since society would likely favor low-cost measures such as increased NATURE CLIMATE CHANGE 1

2 energy efficiency initially while turning to more complex structural changes such as an electricity grid dominated by intermittent renewables more slowly. Relative to the 2030 baseline, this represents a cut of 62% of CO 2 eq emissions. Achieving this via energy and surface transport only requires a cut of about 77% in those sectors combined emissions. Although there may be some progress in industrial emissions controls, the baseline scenario already includes reductions in CO 2 emissions from production of feedstocks and cement, along with fairly strong HFC controls, and agricultural emissions (especially non-co 2 ) are likely to have quite limited mitigation potential accounting for increasing population and hence demand (2). Combined with increased aircraft transportation, this suggests that only modest additional gains might be achieved via the smaller sectors beyond energy and surface transportation. Hence we conclude that reductions of about 60-80% relative to the baseline are needed in energy and surface transport. Note that the energy reductions of 63% represent a total clean energy penetration of 70%, a high fraction given the need for baseload from non-wind and solar or widespread storage capacity, motivating the larger share of reductions apportioned to transport. The clean energy scenario is consistent with the goal of 80% clean electricity by 2035 put forth by President Obama in his 2011 State of the Union address, but expanded to encompass similar advances in industrial, commercial and residential energy as well. In this scenario, the CO 2 reductions are distributed roughly 2-to-1 between electricity and non-electricity sub-sectors (current emissions are similarly distributed). For comparison, US energy sector sulfur dioxide (SO 2 ) emissions are projected to decrease under RCP8.5 due to pollution controls and a continued shift from coal to natural gas, but the latter also leads to increases in emissions of methane (CH 4 ) (Supplementary Table S1). The clean energy policy greatly enhances the SO 2 reductions, and reverses the rising methane trend. The clean transportation scenario s 75% emissions reductions are beyond mandated 50% gains in fuel economy required under the Tier 3 rules for new light-duty vehicles by 2025 and the greenhouse gas 2 NATURE CLIMATE CHANGE

3 SUPPLEMENTARY INFORMATION regulations to be applied to Heavy Duty Trucks during the 2020s (the EPA s Phase 2 regulations), and are applied to all subsectors. Such a scenario would require behavioral shifts such as greater use of electric vehicles, reduced driving or freight mode-switching. Although greater use of electric vehicles would increase the demands on the energy sector, it would also spur development of improved battery storage, enabling greater use of variable renewable power and hence we assume the increased demand can be compensated for without increased energy-sector emissions. Though ambitious, this transportation scenario is consistent with a technically feasible transformation to a zero fossil fuel economy by mid-century (3) and with a proposed transition to fully electrified transportation using innovative technologies (e.g. in-use charging (4)) and the level of ambition that was found to be required in this sector in a policy report (5). Within this sector, strengthened emissions controls for traditional air pollutants lead to substantial baseline RCP8.5 projected decreases of carbon monoxide (CO), nitrogen oxides (NO x ), and black carbon (BC). The clean transportation policy augments these reductions by roughly 80% for the ozone precursors (CO and NO x ), but provides only ~20% additional reduction for BC as those emissions are already projected to be quite small under the baseline scenario. Though we focused on 2030 in much of the modeling, since a large fraction of the CO 2 lasts for decades to centuries, it makes sense to consider the effect of the CO 2 over longer timescales as well. Doing so, however, requires assumptions about CO 2 emissions past As an illustrative example, we consider the case for which the CO 2 emissions under the scenarios or the baseline simply remain constant after In the baseline RCP8.5 scenario, CO 2 emissions for the US transportation sector peak in 2050 and the 2100 emissions are similar to those in Hence for this sector, the reduction in CO 2 emissions relative to the baseline might be expected to increase from 2030 to 2050, but after 2050 to decrease with time. For the energy sector, CO 2 emissions grow rapidly through 2080 under the RCP8.5 scenario, and hence the reduced CO 2 emissions resulting from an early switch to renewables would likely be even larger if post-2030 NATURE CLIMATE CHANGE 3

4 emissions following the policies remained low despite demand growth (i.e. the 70% fraction of energy from clean, non-co 2 emitting sources could be maintained). Radiative Forcing Global mean forcing due to worldwide emissions under the RCP8.5 scenario is projected to increase by 1690 mw m -2 from 2000 to 2030 (6). In comparison, the reductions achieved under these policies appear small. Given that they stem from policies in only a single sector within a single country, however, the fact that they can reduce forcing by a few percent makes them nonetheless important. In the case of the transportation scenario, there are larger differences between the models for US RF (Figure 1; Supplementary Table S3). In the aerosol microphysics model, the direct aerosol effect as nitrate aerosols are not included and hence the net positive aerosol forcing is weaker, though it still outweighs the CO 2 forcing in earlier years as in the model excluding aerosol microphysics. Given the importance of nitrate in this particular scenario, we favor the mass-based results for this sector. Note that the difference between the scenarios appears to be robust to uncertainties in aerosols, with the stronger positive aerosol forcing in the mass-based model causing a slower transition to national level negative net RF in either scenario, for example. Examining ozone forcing, for the energy policy the US average is fairly similar to the global mean ozone forcing as much of the ozone decrease is due to methane and so is fairly uniformly distributed. For the transportation policy, ozone forcing over the US is more than double the global mean value (-12 vs -5 mw m -2 ) as reductions are driven by decreased CO and NO x and hence are much more localized. Methane and CO 2 forcings are nearly uniform over the globe, of course. 4 NATURE CLIMATE CHANGE

5 SUPPLEMENTARY INFORMATION We also tested linearity by performing ½ magnitude transportation reduction simulations using the model version incorporating mass-based aerosols, and find that both the global and US RF forcing results are quite linear (within 3% when doubled). We note that while the main text discusses the effect of the localized aerosol forcings on surface temperature, the localized forcings (in addition to the quasi-homogeneous CO 2 forcing) will also affect precipitation. Positive sulfate aerosol forcing may increase warm-season average US rainfall (7), as CO 2 forcing tends to do, although cooling in the tropical Atlantic (Fig. 3) could reduce the frequency of hurricane formation. Such responses merit further study. Health and Economic Impacts Within PM2.5-related benefits, half to 2/3 come from reduced cardiovascular disease, with the remainder roughly equally split between avoided lung cancers and respiratory disease/infection (in the CRF base and CRF high cases, the CRF low case does not include respiratory diseases). Impacts are fairly linear comparing the ½ magnitude transportation scenario, which is not surprising given that US concentrations are generally in a linear regime (8), with differences within 7% for all cases except the power-law CRF low for which differences are ~13%. Economic valuation uses a variety of economic discount rates spanning the range used by the US government (9) to that used by the UK Treasury (10). Note that the SCAR metric incorporates health impacts using the CRF base only and does not probe the influence of the alternative representations of the CRF examined here. As noted in the main text, the SCAR valuation uses globally representative values which are ~20-25% lower than specific values for health impacts calculated from these experiments. Hence the total valuation would be $45B and $25B greater for energy and transportation, respectively, using the specific NATURE CLIMATE CHANGE 5

6 health monetization from this study within the SCAR (worldwide health benefits would increase from $180B to $225B and from $80B to $105B, respectively). Benefits due to improved air quality are evaluated here incorporating only the valuation of avoided premature deaths related to PM2.5 and ozone. Additional reductions in premature deaths could occur if, for example, a portion of the transportation emissions reductions were achieved via increased active transport (11). Non-mortality benefits would also accrue owing to reduced chronic illnesses, reduced missed school days, reduction of the detrimental effect of poor air quality on cognitive functioning, etc. As with premature deaths, these can be valued based on willingness-to-pay, but typically values (for those for which valuations have been established) are greatly outweighed by the valuation associated with averted premature death (e.g. (12)). Aside from willingness-to-pay, there are also analyses examining the effect of degraded air quality on labor supply and leisure time available to workers, which find additional damages on the order of $2-10 per ton CO 2 for co-benefits related to CO 2 from the energy sector (13) that would increase the valuations reported here. Additional Comparison with Prior Health and Valuation Studies Studies of the transformation of the US economy under low carbon policies have often used detailed models of the energy and transportation systems to generate scenarios. These may be more realistic than the scenarios used here, but results can be difficult to interpret as many factors change simultaneously. In companion studies with the same modeling setup, we have explored the impact of energy sector changes derived with a detailed energy systems model. We find climate impacts that are comparable to the sum of our sectoral results scaled by the relative CO 2 reductions when examining low carbon scenarios, and health impacts that are comparable to these results scaling by the relative SO 2 changes when examining the effects of air quality regulations. This suggests that the idealized scenarios examined here are useful for understanding the role of individual sectors under potential near-term policies as well as for the more ambitious changes envisioned in these scenarios. We also note that the transformations explored in our 6 6 NATURE CLIMATE CHANGE

7 SUPPLEMENTARY INFORMATION energy policy scenario are so large that the question of how the mix of electricity generation or vehicles changes is less important than for smaller interventions as most fossil fuel-based electricity or transportation is removed. Thus although our scenarios are idealized, they appear to provide useful insight into the benefits that could be achieved via economy-wide transitions. As noted in the main text, comparison with prior studies of emissions reduction policies for the US energy and transport sectors is complicated by the differences in the derived changes for the various relevant pollutants under different scenarios. A study of shifts in the US energy and transportation sector using detailed systems models recently focused on valuation of health co-benefits (14). Under their clean energy scenario for 2030, relative to their 2012 baseline they found comparable decreases in SO 2 emissions (-2.8 Tg) to those used in our study (-2.4 Tg) and slightly less than twice the NO x decreases (-0.98 versus Tg). Their reductions in premature deaths, which were, like ours, estimated using multiple CRFs (two for PM2.5), were centered around 30,000 (spanning ~12,000-61,000). These are larger than our 20,000 (10,000 90,000), though well within uncertainty ranges except for uppermost end of our range that comes from our use of a highly non-linear PM2.5 CRF, and appear consistent with the greater NO x changes. In their transportation scenario, NO x emissions decreased ~3 times more than in our scenario whereas CO reductions were similar. Their reduction in premature deaths was ~35,000 (15,000-70,000) versus our ~14,000 ( ,000), so again seemingly consistent with their decrease in NO x being greater (and the large impact of NO x on nitrate aerosols in our model). Their valuation of human health benefits was ~200$B for each sector, so comparable to our results but apparently using a lower per capita VSL (which may be a present-day value whereas we used a 2030 value). Changes in CO 2 emissions associated with those scenarios were not presented, however. Another study looked at health impacts of low carbon energy sector policies (15). Their lowest carbon scenario for 2020 had annual emissions changes of -816 Mt CO 2, Mt SO 2, and Mt NO x relative to 7 NATURE CLIMATE CHANGE 7

8 their 2020 baseline. These values are very close to 1/6 the sum of our clean energy and transportation scenarios, which would give Mt CO 2, Mt SO 2, and Mt NO x. Their result for avoided premature deaths (~3200 yr -1 ) falls within the range for 1/6 th of our estimates ( yr -1 ). Driscoll et al. (2015) also examined an EPA Clean Power Plan scenario for 2020 which had ~1/4-1/5 th our clean energy emissions reductions for air-quality related pollutants and ~1/6 for CO 2, suggesting that the power sources with the greatest traditional pollutant emissions might be controlled first. Hence the air quality benefits of clean energy policies could be greater than those estimated here, at least initially, although that is obviously scenario dependent. Emissions changes in their EPA Clean Power Plan scenario for 2020 were Mt SO 2 yr -1, Mt NO x yr -1, -251 Mt CO 2 yr -1 (no estimate of avoided premature deaths was given for this case). The US EPA has also reported analysis of the proposed Clean Power Plan, stating that it would reduce CO 2 emissions by 730 Tg (16). Our CO 2 cuts spanning the broader energy sector are 1491 Tg relative to current emissions, so almost exactly double. The resulting benefits can be compared, as the EPA reports $55-95B in air-quality and climate related benefits, whereas we find $ B (which becomes $ B dividing by 2). Hence our values are 4-9x greater, consistent with the larger valuation of the broader SCAR metric relative to the traditional SCC. This is primarily due to the inclusion within the SCAR of climate impacts of non-co 2 emissions and to the valuation of climate-related deaths consistently with air-quality related deaths and incorporating all impacts quantified by the World Health Organization rather than the smaller and narrower health valuation within the traditional SCC (17). Implementation Putting into place emissions reductions along the lines of those discussed here would be a tremendous challenge, not least owing to the disbenefits that accrue to particular groups (e.g. the fossil-fuel industry, some automakers). For example, within the transportation sector, electric vehicles are already commercially produced and their use as short-range passenger vehicles is growing rapidly. Transformation to an electrified 8 NATURE CLIMATE CHANGE

9 SUPPLEMENTARY INFORMATION vehicle fleet would be difficult for some sectors, but is quite feasible for most of the country s vehicles and would primarily require additional charging infrastructure (perhaps even in-use charging (4)) and renewable energy generation (perhaps with continuation of tax incentives initially to encourage adoption in the absence of a damage-recovery fee applied to gasoline vehicles that incorporates their actual societal cost (17)). Yet vehicle manufacturers that do not produce electric vehicles would stand to lose. At present, within both sectors, those businesses accruing disbenefits are much larger than those businesses that would accrue benefits. These misaligned incentives create formidable implementation barriers at the national scale. We note that misaligned incentives are unlikely to play a similarly important role at the international scale, as although a nation eschewing reductions in its own emissions could temporarily reap an outsized climate benefit via avoiding local disbenefits and free-riding on the remote benefits of others actions, they would pay a heavy price in local air quality damages. NATURE CLIMATE CHANGE 9

10 Energy Policy Net Forcing, 2050 Microphysics Model Net Forcing, 2030 Microphysics Model Cloud-Aerosol Forcing Direct Aerosol Forcing mw m Supplementary Figure S1. Radiative forcing in the aerosol microphysics model. As in Fig. 1 except for using results from the model version incorporating aerosol microphysics for the clean energy policy. Note that that model does not include nitrate aerosols, which are important in the clean transportation scenario (hence the latter is omitted here). 10 NATURE CLIMATE CHANGE

11 SUPPLEMENTARY INFORMATION Energy Policy Transportation Policy Ozone ppbv Avoided premature deaths per million yr -1 Supplementary Figure S2. Change in maximum 6-month average of the 1-hr daily surface ozone maximum and ozone-related mortalities. Values are presented for the two policies from the massbased model (results are very similar for the microphysics model), with health impacts calculated using the concentration-response function associated with long-term ozone exposure. NATURE CLIMATE CHANGE 11

12 Supplementary Table S1. US emissions changes. There are also changes in emissions of ammonia and volatile organic compounds, which are not shown as these are very small. Baseline RCP relative to 2010 Policy Scenarios for 2030 relative to baseline 2030 Energy sector Transportation sector Clean energy Clean transportation SO 2 (Tg SO 2 /yr) CH 4 (Tg CH 4 /yr) CO (Tg CO/yr) NO x (Tg N/yr) OC (Tg C/yr) BC (Tg C/yr) CO 2 (Tg CO 2 /yr) Supplementary Table S2. Global mean radiative forcing (mw m -2 ) in Statistical significance is ~0.3 mw m -2. O 3 CH 4 SO 4 BC OC NO 3 Aerosol Aerosol- CO 2 Net Direct Cloud Aerosol mass-based model Energy policy Transportation policy Aerosol microphysics model Energy policy N/A Transportation policy N/A NATURE CLIMATE CHANGE

13 SUPPLEMENTARY INFORMATION Supplementary Table S3. Radiative forcing (mw m -2 ) over the US in Statistical significance is 1 mw m -2 for ozone and 2 mw m -2 for aerosol forcings. O 3 CH 4 Aerosol Aerosol- CO 2 Net Direct Cloud Aerosol mass-based model Energy policy Transportation policy Aerosol microphysics model Energy policy Transportation policy NATURE CLIMATE CHANGE 13

14 Additional References 1. K. Riahi et al., RCP 8.5-A scenario of comparatively high greenhouse gas emissions. Climatic Change 109, (2011). 2. D. Gernaat et al., Understanding the contribution of non-carbon dioxide gases in deep mitigation scenarios. Global Environmental Change-Human and Policy Dimensions 33, (2015). 3. M. A. Delucchi, M. Z. Jacobson, Providing all Global Energy with Wind, Water, and Solar Power, Part II: Reliability, System and Transmission Costs, and Policies. Energy Policy 39, (2011). 4. K. Heaslip, K. C. Womack, J. Muhs, Automated Electric Transportation: A Way to Meet American's Critical Issues. Leadership Manage. Eng. 11, (2011). 5. J. H. Williams et al., "Pathways to deep decarbonization in the United States. The U.S. report of the Deep Decarbonization Pathways Project of the Sustainable Development Solutions Network and the Institute for Sustainable Development and International Relation," (2014). 6. D. T. Shindell et al., Radiative forcing in the ACCMIP historical and future climate simulations. Atmospheric Chemistry and Physics 13, (2013). 7. D. T. Shindell, A. Voulgarakis, G. Faluvegi, G. Milly, Precipitation response to regional radiative forcing. Atmos. Chem. Phys. 12, (2012). 8. J. Schwartz, B. Coull, F. Laden, L. Ryan, The effect of dose and timing of dose on the association between airborne particles and survival. Environmental Health Perspectives 116, (2008). 9. US Government, "Technical Update of the Social Cost of Carbon for Regulatory Impact Analysis Under Executive Order 12866," (Interagency Working Group on Social Cost of Carbon, 2013). 10. N. Stern, Stern review on the economics of climate change. (UK Treasury, London, 2006). 11. J. Woodcock et al., Public health benefits of strategies to reduce greenhouse-gas emissions: urban land transport. Lancet 374, (2009). 12. N. Fann et al., The geographic distribution and economic value of climate change-related ozone health impacts in the United States in Journal of the Air & Waste Management Association 65, (2015). 13. R. K. Saari, N. E. Selin, S. Rausch, T. M. Thompson, A self-consistent method to assess air quality co-benefits from U.S. climate policies. Journal of the Air & Waste Management Association 65, (2015). 14. T. M. Thompson, S. Rausch, R. K. Saari, N. E. Selin, A systems approach to evaluating the air quality co-benefits of US carbon policies. Nature Climate Change 4, (2014). 15. C. T. Driscoll et al., US power plant carbon standards and clean air and health co-benefits. Nature Climate Change 5, (2015). 16. US EPA, in Clean Power Plan Fact Sheet. (2014), vol D. Shindell, The Social Cost of Atmospheric Release. Clim. Change 130, (2015). 14 NATURE CLIMATE CHANGE