CLIM-D-16-00553 – yellow highlights = changes for Galley
Climatic Change Electronic Supplementary Material
CLIM-D-16-00553 DOI 10.1007/s10584-017-1978-0 ESM-1 May 2017
The rise in global atmospheric CO2, surface temperature and sea level from emissions traced to major carbon producers
B. Ekwurzel,1 J. Boneham,2 M. W. Dalton,2,3 R. Heede,4 R. J. Mera,1,5 M. R. Allen,2 P. C. Frumhoff,6
1Union of Concerned Scientists, Washington DC, 20006, USA
2University of Oxford, Oxford, OX1 3QY, U
3Present Address: London, UK
4Climate Accountability Institute, Snowmass, CO, 81654, USA
5Present Address: North Carolina State University, Raleigh, NC, USA
3Union of Concerned Scientists, Cambridge, MA, 02238, USA
1 Climate Model
1.1 Impulse response function
The simple climate model is based on the approach presented in IPCC AR5 supplementary material for chapter eight of working group one (Myhre et al., 2013a). The impulse response models are largely based on equations featured in the IPCC AR5 that combine the shorter timescale response processes of the atmosphere, land surface, and upper ocean with the longer timescale response processes of the deep ocean (Myhre et al. 2013a; Joos et al. 2013; Otto et al. 2013; Millar et al. 2015). Climate sensitivity is among the significant sources of climate system uncertainty (Forest et al. 2002). Typical impulse response model simulations are varied over the ranges for transient climate response and equilibrium climate sensitivity co-varying according to the inherent ratio relationship between the two (Millar et al. 2015).
In this study we use the modified impulse response equations that incorporate the response of atmospheric CO2 to temperature change and CO2 accumulation (Millar et al. 2016). This is primarily due to the reduced carbon uptake by terrestrial and marine carbon reservoirs with increasing temperature (Friedlingstein et al. 2006; Millar et al. 2016). It is an impulse response of CO2 to emissions with a “4-time constant” for the carbon reservoirs and a temperature response to radiative forcing with a “2-time constant” that has been expanded to allow a single scaling factor to the four carbon reservoir times (Millar et al. 2016). First the AR5 model will be reviewed below followed by the expansion to allow for a responsive carbon cycle.
The impulse response function (IRFx) is the time-dependent abundance of gas x caused by the additional emission of one kilogram (kg) of gas x at time 0 (Joos et al., 2013). The IRFx(t) is the fraction of concentration enhancement due to the added emission pulse remaining in the atmosphere at time t:
IRFxt=ax,0+i=1nax,i∙exp-tτx,i,
(1)
where ax,i is the fraction associated with a certain timescale τx,i and ax,0 is the fraction of emissions that remains permanently in the atmosphere (Joos et al. 2013). The multi-model mean for IRFCO2 yield the four coefficients in table 5 in Joos et al (2013) and table 8.SM.10 in Myhre et al (2013). To these four time-constants a single scaling factor is applied to update the concentrations in the four carbon “pools”:
dRidt=aiE-Riατi ,i=1,4 (2)
where E are annual emissions (ppm/year)(Millar et al. 2016). Atmospheric CO2 concentrations are given by:
C=C0+iRi, (3)
and radiative forcing by:
F=F2xln2lnCC0+ Fext , (4)
where C0 is the pre-industrial CO2 concentration, F2x the forcing due to CO2 doubling, and Fext the non-CO2 forcing (Millar et al. 2016). Global mean surface temperature (GMST) change from average are:
dTjdt=cjF-Tjdj ; T=jTj ;j=1,2 , (5)
with coefficients ai, dj and τi as given in AR5 Chapter 8, tables 8.SM.9 and 8.SM.10 (Boucher and Reddy 2008; Myhre et al. 2013a; Millar et al. 2016). The only difference between this model and that used for metric calculations in AR5 is the cj are set to give an equilibrium climate sensitivity (ECS) of 2.75K and transient climate response (TCR) of 1.6K, and the state-dependent coefficient α (Millar et al. 2016). See the best estimate settings in table ESM-1.
To identify a suitable state-dependence, the 100-year integrated impulse response function, iIRF100 was implemented with other coefficients fixed, this is a non-linear function of α (Millar et al. 2016):
iIRF100=iαaiτi 1-exp-100ατi . (6)
The iIRF100 is assumed to be a separable function of accumulated perturbation carbon stock in the land and ocean, Cacc=tE-(C-Co), and the GMST change from pre-industrial, T (Millar et al. 2016). The following functional form gives a good approximation to the behavior of more complex models (Millar et al. 2016):
iIRF100=r0+rCCacc+rTT . (7)
The following parameter values appear to be acceptable as the best estimate case (Millar et al. 2016): r0=35years ; rC=0.02yearsGtCand rT=4.5years/K ; with ECS=2.7K and TCR=1.6K give iIRF100 of 53 years for a 100 GtC pulse released against a 389ppm background CO2 concentration following a historical build-up, consistent with the Joos et al., (2013) central estimate (Millar et al. 2016).
For a prognostic model, the computed iIRF100 at each time-step uses Cp and T from the previous time-step, converts to α and applies it to the carbon cycle equations (Millar et al. 2016). This means the iIRF100 is only exactly reproduced under constant background conditions with infinitesimal perturbations (Millar et al. 2016).
Model simulations for full historical forcing and best estimate parameters closely matched historical observations (Fig. 1 and 2, ESM Table 5 and 6) (Millar et al. 2016). Sensitivity tests were simulated using the high and low range (ESM Table 2) of equilibrium climate sensitivity and related transient climate response (Myhre et al. 2013b; Millar et al. 2015).
Table 1 Carbon scaling parameters
Carbon Settings / Low / High / BestEmissions scaling factor / 0.86 / 1.14 / 1
Decay time scaling factor / 43.8 / 30.8 / 35
Table 2. Thermal parameters
Thermal Settings / Low / High / BestTCR / 1 / 2.5 / 1.6
ECS / 1.5 / 4.5 / 2.75
Adjustment to non-GHG forcing / -1.1 / 0.65 / 0
1.2 Forcers in the climate model
The annual average global mean radiative forcing as well as emissions data was provided by MAGICC version 6.3.09 RCP8.5 for 2005-2010 and including the historical data provided before 2005 (Meinshausen et al. 2011). Carbon dioxide forcing is based on the impulse response approach described in section 1.1 drawing from the combined historical emissions for fossil and industrial CO2 (fossil, cement, gas flaring and bunker fuels) plus land use related CO2 emissions. The other greenhouse gas forcing were taken directly from the radiative forcing data for total greenhouse gas forcing (CO2, CH4, N2O, HFCs, PFCs, SF6, and Montreal Protocol gases) minus the CO2 forcing data. The other anthropogenic forcing were obtained from the total anthropogenic forcing data minus the total greenhouse gas forcing (CO2, CH4, N2O, HFCs, PFCs, SF6, and Montreal Protocol gases) data. The natural forcing was derived from the annual mean volcanic stratospheric aerosol forcing data plus the solar irradiance forcing data. The user can choose to run the impulse response model without the total annual historical fossil fuel aerosols. This choice takes the total forcing data (i.e. CO2 forcing plus other greenhouse gas forcing plus other anthropogenic forcing plus natural forcing described above) minus total historical fossil fuel aerosols data (i.e. direct fossil fuel aerosol (organic carbon) minus direct fossil fuel aerosol (black carbon) minus direct sulphate aerosol minus direct nitrate aerosol) (Meinshausen et al. 2011).
The methane (CH4) emissions traced to carbon producers (Heede 2014) are incorporated into the model in the following way. The model uses a lifetime for CH4 of around 9 years. This is in agreement with the IPCC AR5 total CH4 lifetime in the atmosphere (9.25 ± 0.6 years) based on different lifetimes with respect to tropospheric hydroxyl radical (OH), stratospheric loss, chlorine loss, and bacterial uptake in soils (Myhre et al 2013b). The reference total annual CH4 emissions are from MAGICC (see above). Depending on the user selection of which carbon producers to remove emissions over a chosen start date through 2010, each year will have the reference MAGICC (see above) total methane emissions or the reference emission minus the carbon producers for that year’s annual CH4 emissions. The model starts in 1765 with the atmospheric concentration anomaly set to zero and the atmospheric concentration set to the pre-industrial value. The following year the average of that year and the prior year’s emissions are added to the prior year’s concentration anomaly (units of ppb). The exponential decay term is calculated for the prior year’s concentration anomaly using the CH4 lifetime and subtracted from the current year’s concentration anomaly. The total concentration anomaly for the year is added to the pre-industrial atmospheric CH4 concentration (722 ppb) to get the current year’s atmospheric concentration. This continues for each year through 2010 to create the atmospheric CH4 concentration plot showing the MAGICC reference case compared with cases for selected carbon producers emissions removed. To see the relative contributions the ‘Dashboard’ to choose model simulations includes a table of Heede (2014) data for CH4 (MtCO2eq) and CO2 (MtCO2) emissions traced back to carbon producers. Finally, the Radiative Efficiency for CH4 (0.000417) was multiplied by the annually calculated total concentration anomaly (described above) to obtain the annual CH4 forcing in watts per square meter (W m-2) and added to the other annual forcings (W m-2) in the GMST calculations.
1.3 Global Sea Level
The rate of global sea level (GSL) change is based on the model output for the global mean surface temperature using the equations in Kopp et al., 2016 and the parameters in ESM table 3. Parameter a is the GSL rate sensitivity to a deviation of temperature from an equilibrium temperature andτ1 is the timescale the temperature relaxes to the equilibrium temperature (Kopp et al. 2016).
Table 3. Sea Level parameters (Kopp et al., 2016)
Parameter / Low / High / Besta / 0.34 / 0.7 / 0.47
τ1 / 64 / 203 / 102
c(2000) / 0.001 / 0.006 / 0.003
c timescale / 1124 / 16155 / 3392
1.4 Sensitivity Tests
Fig. 1 Model simulations for the reference case (i.e. no removal of emissions traced to carbon producers) compared with historical observations from the U.S. Environmental Protection Agency Climate Indicators(US EPA). Abbreviations: IA1 = full forcing including total historical fossil aerosols; IA0 = full forcing minus total historical fossil aerosols; Best, High, Low are parameters that correspond to values in ESM tables 1, 2 and 3. Radiative forcing was reduced after major volcanic eruptions (Grimsvotn, Iceland (1783), Tambora, Sumbawa, Indonesia (1815), Cosiguina, Nicaragua (1835), Askja, Iceland (1875), Krakatau, Indonesia (1883), Okataina [Tarawera], North Island, New Zealand (1886), Santa Maria, Guatemala ( 1902), Ksudach, Kamchatka, Russia (1907), Novarupta [Katmai], Alaska, US (1912), Agung, Bali, Indonesia (1963), Mount St. Helens, Washington, US (1980), El Chichón, Chiapas, Mexico (1982), and Mount Pinatubo, Luzon Philippines (1991); (Robock 2000)).
Fig. 2 Global mean surface temperature (°C) after removing annual emissions traced to 90 major industrial carbon producers starting in 1880. Model simulations implemented with the best estimate parameters and full historical forcing. Dark blue line is the reference case (i.e. no removal of emissions traced to carbon producers). First 6 lines below the reference case are after removal of emissions traced to Chevron, ExxonMobil, Saudi Aramco, BP, Gazprom and Shell. The first two large gaps after removal of emissions traced to investor-owned or majority state-owned carbon producers are the Former Soviet Union and China (coal & cement only). Historical temperature is the semi-transparent line with triangle markers.
Table 4. Contribution from emissions (1880-2010; 1980-2010) traced to individual carbon producers to increase in atmospheric CO2 1880-2010 (ESM-2 available online)
Table 5. Contribution from emissions (1880-2010; 1980-2010) traced to individual carbon producers to increase in global mean surface temperature (GMST) 1880-2010 (ESM-2 available online)
Table 6. Contribution from emissions (1880-2010; 1980-2010) traced to individual carbon producers to increase in global = sea level (GSL) 1880-2010 (ESM-2 available online)
File ESM-3 (available online) is the global energy-balance coupled-climate-carbon-cycle model used to assess the change in atmospheric CO2 concentrations, atmospheric CH4 concentrations, radiative forcing, GMST and GSL resulting from emissions traced to 90 major carbon producers.
References Electronic Supplementary Material
Boucher O, Reddy MS (2008) Climate trade-off between black carbon and carbon dioxide emissions. Energy Policy 36:193–200. doi: 10.1016/j.enpol.2007.08.039
Forest CE, Stone PH, Sokolov AP, et al (2002) Quantifying Uncertainties in Climate System Properties with the Use of Recent Climate Observations. Science 295:113–117. doi: 10.1126/science.1064419
Friedlingstein P, Cox P, Betts R, et al (2006) Climate–Carbon Cycle Feedback Analysis: Results from the C4MIP Model Intercomparison. J Climate 19:3337–3353. doi: 10.1175/JCLI3800.1
Heede R (2014) Tracing anthropogenic carbon dioxide and methane emissions to fossil fuel and cement producers, 1854–2010. Climatic Change 122:229–241.
Joos F, Roth R, Fuglestvedt JS, et al (2013) Carbon dioxide and climate impulse response functions for the computation of greenhouse gas metrics: a multi-model analysis. Atmos Chem Phys 13:2793–2825. doi: 10.5194/acp-13-2793-2013
Kopp RE, Kemp AC, Bittermann K, et al (2016) Temperature-driven global sea-level variability in the Common Era. PNAS 113:E1434–E1441. doi: 10.1073/pnas.1517056113
Meinshausen M, Smith SJ, Calvin K, et al (2011) The RCP greenhouse gas concentrations and their extensions from 1765 to 2300. Climatic Change 109:213–241. doi: 10.1007/s10584-011-0156-z
Millar RJ, Nicholls ZR, Friedlingstein P, Allen MR (2016) A modified impulse-response representation of the global response to carbon dioxide emissions. Atmospheric Chemistry and Physics Discussions 1–20. doi: 10.5194/acp-2016-405
Millar RJ, Otto A, Forster PM, et al (2015) Model structure in observational constraints on transient climate response. Climatic Change 131:199–211. doi: DOI 10.1007/s10584-015-1384-4
Myhre G, Shindell D, Bréon FM, et al (2013a) Anthropogenic and Natural Radiative Forcing Supplementary Material. In: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change In: [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)].