Electronic Supplementary Material for
The Long-Term Policy Context for Solar Radiation Management
Steven J. Smith[1] and Philip J. Rasch[2]
CONTENTS
ESM 1 –Methods 1
MAGICC Climate Model 1
ESM 2 –GCAM Scenarios 3
Historical Emissions 3
Future Emissions and GCAM Reference Scenario 3
GCAM Climate Policy Scenarios: RCP4.5-stab and GCAM-PD 4
4.5Stab -> Peak and Decline (PD) Scenarios 5
ESM 3 –Example SRM Application 6
ESM References 7
ESM 1 –Methods
MAGICC Climate Model
The MAGICC climate model includes a comprehensive suite of forcing agents including greenhouse gases, stratospheric ozone depleting substances, aerosols, and tropospheric ozone precursor compounds. The assumed historical values for these forcings are shown in Table ESM-1.
Greenhouse gas concentrations, radiative forcing, and global climate changes in this paper are modeled with the GCAM 3.1 version of the MAGICC 5.3 model as used in the IPCC 4th assessment report. MAGICC is a simple mechanistic model that includes the ocean and terrestrial carbon-cycle (including CO2 and temperature feedbacks), parameterized representations of atmospheric chemistry (tropospheric ozone and methane oxidation), a comprehensive suite of radiative forcing agents, differential land-ocean climate sensitivity, and an upwelling-diffusion representation of ocean heat transport. Using input data on anthropogenic emissions (starting in 1990) and historical concentrations and aerosol emissions, MAGICC estimates changes in top of the atmosphere radiative forcing (starting in 1765), and the subsequent changes in global temperature change in four boxes (land/ocean north and southern hemisphere).
The GCAM model, including the MAGICC component, is described on-line, and also is available as open source software at: http://www.globalchange.umd.edu/models/gcam/. The version of the MAGICC model used here is version 5.3, re-coded in C++. The major change in this version is, unlike the original version of MAGICC 5.3, black carbon and organic carbon forcing are assumed to be proportional to anthropogenic emissions of BC and OC and are calibrated to central values of forcing per unit emission from Bond et al. (2011). Base year radiative forcing assumptions are shown in Table ESM-1. Note that future forcing for land albedo, mineral dust, and nitrate aerosols are assumed constant since reduced form representations of these forcing agents have not been incorporated into this version of MAGICC.
Forcing Agent / 2000 / 2005Carbon Dioxide (CO2) / 1.52 / 1.65
Methane (CH4) / 0.49 / 0.51
Nitrous Oxide (N2O) / 0.15 / 0.17
Fluorinated Gases / 0.02 / 0.02
Trop. Ozone (inc. CH4 component) / 0.33 / 0.36
Montreal Protocol gases / 0.34 / 0.35
Direct Sulfate / -0.33 / -0.36
Direct Nitrates / -0.10 / -0.10
Cloud indirect / -0.62 / -0.65
Direct Organic Carbon (OC) / -0.07 / -0.08
Direct Black Carbon (BC, inc. on Snow) / 0.35 / 0.38
Stratospheric Ozone / -0.22 / -0.20
Stratospheric H2O forcing f/ CH4 oxidization / 0.02 / 0.03
Land albedo / -0.20 / -0.20
Mineral dust / -0.10 / -0.10
Total Anthropogenic Forcing / 1.59 / 1.79
Table ESM-1. Year 2000 2005 anthropogenic forcing in GCAM 3.1.
Radiative forcing as used here follows the definition used in the RCP scenarios and includes: all well-mixed greenhouse gases, tropospheric and stratospheric ozone changes, stratospheric water vapor feedback, sulfate aerosols, cloud indirect effect, black carbon (BC), and organic carbon (OC). Land albedo and nitrate aerosols were not included in this total. While the MAGICC model used here includes these two components, land-albedo and nitrate aerosol forcing is assumed to be constant over the 21st century at -0.4 W/m2. Note that equilibrium temperature change would, therefore, be proportional to the climate sensitivity times the RCP forcing minus 0.4 W/m2. Note also that forcing in the RCP4.5-stab scenario in 2100 as used here is 4.7 W/m2, which is slightly higher than the nominal value from the original RCP4.5-stab scenario calculation (Thomson et al. 2011) due largely to revised historical calibration and the improved representation of BC and OC forcing.
Figure ESM-1 shows the historical temperature change as simulated by the version of MAGICC as used in this paper (solid lines) as well as a model run from the same model that also includes estimates of solar and volcanic forcings (dashed line). The model results used here do not include estimates of solar and volcanic forcings, which have a negligible impact on future results (as shown in the figure). Solar and volcanic forcings, however do have a substantial impact on historical results, and are necessary to reproduce the historical pattern. The historical MAGICC results capture the primary features of the instrumental record (for example, Brohan et al. 2006).
Figure ESM-1. Historical global-mean temperature change as simulated by the version of MAGICC as used in this paper (solid lines) as well as a model run that also includes estimates of solar and volcanic forcings (dashed line).
ESM 2 –GCAM Scenarios
Historical Emissions
Historical emissions for all substances use estimates from the GCAM 3.1 model up to 2005.[3] These estimates are derived from country level inventory data (UNFCCC emissions reports, UK DEFRA, Environment Canada, Eyring et al. (2009) and Buhaug et al. (2009), US EPA), with values from the EDGAR 4.2 database where country-level inventories were not available, Meinshausen et al. (2011) for fluorinated gases, and Smith et al. (2011a) for SO2. Carbon dioxide emissions from Peters et al. (2012)[4] are used out to 2010. Emissions were linearly interpolated between the last inventory year (2005, except 2010 for CO2) and the GCAM reference scenario 2020 value. Emissions are identical for all scenarios until 2015 using these extrapolated values. Emissions in the climate policy scenarios, begin to diverge from the reference case in 2016.
Future Emissions and GCAM Reference Scenario
Emissions for all substances starting in 2020 for the GCAM reference, RCP4.5-stab, and GCAM-PD scenarios are taken from Thomson et al (2011) and are available on-line.[5] Compete emission input files for the scenarios used in this analysis are also provided as supplementary material to this article. Emissions for not only greenhouse gases, but also aerosol and tropospheric ozone precursor compounds are, therefore, consistent across all scenarios, and show consistent changes between the scenarios. Sulfur dioxide emissions, for example, are lower in the RCP4.5-stab and GCAM-PD scenarios because climate policies reduce the use of fossil fuels that are the primary source of these emissions.
The GCAM reference scenario begins with the scenario described by Clarke et al (2007), using an updated land-use model (Wise et al 2009), non-CO2 emissions and emissions controls implemented as described by Smith and Wigley (2006) and Smith et al. (2011b), and updated year 2000 inventory information developed for the RCP scenario process (Lamarque et al 2010).
The GCAM reference scenario represents a world where global incomes increase substantially over the 21st century, with global GDP increasing by a factor of ten (Clarke et al 2007). Population increases by 40%, a slight decline after a peak of 9 billion after mid-century, and primary energy use triples. GDP per capita increases over the century such that most world regions are near current OECD levels by the end of the century. Further details on the development of the pollutant emission scenarios for the GCAM reference scenario are given in Smith et al. (2011b).
Radiative forcing in the GCAM reference scenario reaches 7.1 W/m2 in 2100, which is lower than the RCP8.5 scenario (which represents the high end of reference scenarios), but higher than the RCP6.0 scenario (which is representative of either a modest climate policy scenario, or a mid- to lower-end reference scenario). Again, for consistency with the RCP nomenclature, note that forcing figures here follow the RCP convention, which excludes the forcing components enumerated above.
Even higher emission reference scenarios, such as the RCP8.5 scenario, were not included in the analysis here because temperature change in such scenarios will always exceed a 2°C climate target under the cases considered here.
GCAM Climate Policy Scenarios: RCP4.5-stab and GCAM-PD
The GCAM RCP4.5-stab and GCAM-PD (labeled GCAM2.6 in Thomson et al. 2011) scenarios are climate policy scenarios where a global carbon price is applied in order to reduce global radiative forcing to a specified target level. The carbon price is applied to all net carbon emissions and also to all carbon in terrestrial carbon stocks. The RCP4.5-stab scenario was designed to stabilize radiative forcing at 4.5 W/m2 in 2100. In the RCP4.5-stab scenario the carbon price carbon price increases at an annual rate of 5% per year and is approximately constant once stabilization is achieved. In the GCAM-PD scenario the carbon price increases throughout the century, resulting in a “peak and decline” radiative forcing profile whereby radiative forcing at the end of the century is lower than the peak value that occurs near mid-century. Global-mean temperatures also are decreasing by the end of the century, although not at the same rate as the forcing decline due to ocean thermal inertia.
The GCAM-PD scenario, as with other PD scenarios in the literature (Azar et al. 2006, Riahi et al. 2007, van Vuuren, et al. 2007, Calvin et al. 2009, van Vuuren et al. 2011) uses net negative CO2 emissions in order to achieve the low end of century forcing target specified for this scenario. These negative emissions are supplied, for the most part, by capturing the carbon contained in biomass fuel and storing that carbon in geologic reservoirs (also called BECCS, Biomass Energy with Carbon Capture and Storage), although land-use policies, such as re-forestation, also play a role.
Carbon dioxide Capture and Storage (CCS) technologies for fossil fuels are assumed to be deployed in the future in the GCAM scenarios, as is the case in many climate policy scenarios from other integrated assessment models. CCS technologies can substantially lower the cost of meeting climate targets (Edmonds et al. 2004, Edmonds et al. 2007, Global CCS Institute 2011a, IPCC 2005), which provides a substantial incentive for the further development and deployment of these technologies (an incentive that, however, only exists under the implementation of a policy to limit carbon dioxide emissions). Once CCS technologies are assumed to be available, it is consistent to also assume that biomass with CCS technologies are also available. This means that net negative emissions from the power (or refining) sectors are also possible (Luckow et al. 2010), which also lowers climate policy costs, as these negative emissions can offset emissions from sectors such as transportation (particularly freight transport and air travel), which are more difficult to reduce.
The assumption of net negative emission is, therefore, a natural consequence of the assumption that CCS technologies are available for widespread deployment in the future. This has become a common feature of future scenarios as developed by energy-economic models. These negative emissions provide additional flexibility as compared to analysis, based on physical models, of future options that do not consider the possibility of net negative emissions (Matthews & Caldeira 2008, Vaughan et al. 2009).
Underground carbon dioxide injection is now a well-developed technology that has been used for more than 30 years at commercial scales to enhance oil recovery, particularly in West Texas (Dooley et al. 2009, Godec et al. 2011, Meyer 2007, Moritis 2010, Orr and Taber, 1984). There are also four large commercial end-to-end CCS facilities in operation around the world that store anthropogenic CO2 in deep geologic reservoirs explicitly for the purpose of reducing greenhouse gas emissions (Eiken et al. 2011, Global CCS Institute 2011b, Herzog 2011, Whittaker et al. 2011). Extraction of CO2 from industrial operations is also a commercial technology (Herzog et al., 1997, Herzog 2011, IPCC 2005, Moritis 2010, Rubin et al. 2010, van Bergen et al. 2003).( 2003).
CO2 capture and geologic storage has not, however, been demonstrated at the scale required for climate mitigation (e.g., at the scale of a 1000 GW coal-fired power plant). Uncertainty as to the date at which large scale deployment of CCS technology would be possible is one of many uncertainties that apply to future scenarios, and the very low peak and decline forcing pathways in particular. This is one motivation for developing the illustrative transition scenarios (Table 1, main text, and ESM above) whereby the implementation of a forcing peak and decline pathway is delayed by 1 – 3 decades.
4.5Stab -> Peak and Decline (PD) Scenarios
Three scenarios were developed that represent a transition from the 4.5 radiative forcing stabilization scenario to a peak and decline forcing pathway. There are many possible pathways for such transitions and these scenarios were developed assuming that the absolute rate of emissions decrease is similar to that in the GCAM-PD scenario, or, of 2.5-3 GtC/yr per decade. The context of these scenarios is meeting a nominal 2.0 °C long-term climate change target. These transition scenarios were, therefore, constructed such that global-mean temperature change is similar to the GCAM-PD scenario by 2200. In order to counter the additional carbon dioxide emissions that occur due to following the RCP4.5 stabilization pathway for a longer period of time, additional net negative emissions were assumed to occur into the 21st century of a sufficient magnitude and length of time such that global-mean temperatures (under the central value for climate sensitivity) is similar by 2200 (Figures 1 and 2, main text).
In the three transition scenarios, deforestation and non-CO2 emissions initially follow the RCP4.5-stab trajectory and, for simplicity, these emissions are set equal to the GCAM-PD scenario emissions by 2080 (and interpolated between these two scenarios for intermediate years). This means that aerosol and ozone precursor emissions as well as CH4 emissions decline substantially over the century in the PD and transition scenarios, both in absolute terms and relative to the reference scenario. While climate policy results in reductions in emissions related to short-term forcing agents in these scenarios, no additional actions specifically focused on near-term reductions in these agents were considered.
Note that global temperature change in the Stab->PD_10yr and Stab->PD_20yr scenarios is slightly above 1.9°C in 2300, which means that the degree-year metric is not well defined in these cases.
ESM 3 –Example SRM Application
In the example SRM experiments a generic version of Solar Radiation Management (SRM) was implemented as a top of the atmosphere reduction in radiative forcing. Specifically, OC emissions were changed, but the manner in which this change was implemented has a negligible impact on the results as the only impacts of these emissions within the MAGICC model are their impact on direct radiative forcing.