GCAM Shared-Socioeconomic Pathways

Overview of the SSPs

The GCAM model was one of six models used in the quantification of the Shared Socioeconomic Pathways (SSP; Riahi et al., 2017). The SSPs are new reference scenarios for climate change modeling and research (Moss et al. 2010, van Vuuren et al., 2014). That is the five SSPs assume no additional policies and measures explicitly designed to reduce greenhouse gas emissions. However, each of the five reference scenarios can be paired with a set of emissions mitigation assumptions. Thus, when paired with a limit of 2.6, 4.5, or 6.0, the resulting scenario can be used in combination with the corresponding climate model experiments in the CMIP data base. The five SSPs are defined along two different axes: challenges to mitigation and challenges to adaptation (see Figure below).

SSP Definitions
Figure 1: Definitions of SSPs in terms of Challenges to Mitigation and Challenges to Adaptation

Each of the six IAMs produced results for between three and five of the SSPs, with one model identified as the “marker” for each SSP and the others providing estimates of uncertainty. GCAM was the marker model for the SSP4. Marker models, and models providing uncertainty estimates, for all SSPs are listed in the Table below.

Table 1: Overview of the SSPs

Identifier	Descriptor	Marker Model (Institution)	Also computed by
SSP1	Sustainability	IMAGE (PBL)	All
SSP2	Middle-of-the-Road	MESSAGE-GLOBIOM (IIASA)	All
SSP3	Regional Rivalry	AIM/CGE (NIES)	GCAM, IMAGE, MESSAGE-GLOBIOM
SSP4	Inequality	GCAM (PNNL)	AIM/CGE
SSP5	Fossil-fueled Development	REMIND-MAgPIE (PIK)	AIM/CGE, GCAM

Accessing SSP Data

Official SSP data, as planned for use in CMIP6, must be downloaded from the official SSP database.

The GCAM data for the SSP4 and the GCAM representation of all other SSPs (Calvin et al., 2017) are being made available to interested researchers for scenario analysis and comparison.

Implementation in GCAM

To implement the SSPs within GCAM, we use the demographic and economic assumptions developed by Lutz et al. (2015) and Dellink et al. (2015) in combination with technology and policy assumptions derived from the SSP narratives (O’Neill et al. 2015). Table 2 provides a qualitative summary of these parameters; quantitative details are provided in subsequent tables. In general, we have altered parameters related to energy supply (e.g., the capital cost of new power plants, technical change on extraction costs), energy demand (e.g., preferences for traditional bioenergy), agriculture (e.g., agricultural productivity growth), and policy (e.g., start years for carbon policies). To implement the SSP storylines within GCAM, we quantified “high”, “medium”, and “low” for each of the variables listed in Table 2. Those quantifications are provided in Table 3 through Table 8.

Table 2: Qualitative Assumptions across SSPs

Category	Variable	SSP1	SSP2	SSP3	SSP4 (High / Medium / Low Income )	SSP5	Link to Inputs
Socioeconomics	Population in 2100	6.9 billion	9 billion	12.7 billion	0.9 billion / 2.0 billion / 6.4 billion	7.4 billion	Socioeconomics
	GDP per capita in 2100	$46,306	$33,307	$12,092	$123,244 / $30,937 / $7,388	$83,496	Socioeconomics
Fossil Resources (Technological Change/Acceptance)	Coal	Med/Low	Med/Med	High/High	Med/Low / Med/Med / Med/High	High/High	Table 4
	Conventional Gas & Oil	Med/Med	Med/Med	Med/Med	High/Low / High/Low / High/Low	High/High	Table 4
	Unconventional Oil	Low/Med	Med/Med	Med/Med	Med/Low / Med/Low / Med/Low	High/High	Table 4
Electricity (Technology Cost)	Nuclear	High	Med	High	Low /Low / Low	Med	Table 3
	Renewables	Low	Med	High	Low / Low / Low	Med	Table 3
	CCS	High	Med	Med	Low / Low / Low	Low	CCS extraction costs
Fuel Preference	Renewables	High	Med	Med	High / High / High	Med	SSP1
	Traditional Biomass	Low	Low	High	Low / Low / High	Low	Table 5
Energy Demand (Service Demands)	Buildings	Low	Med	Low	High / Med / Low	High	Linked to income
	Transportation	Low and Low	Med	Low	High / Med / Low	High	Links are specific to SSP
	Industry	Low	Med	Low	High / Med / Low	High	Industry and Cement
Agriculture & Land Use	Food Demand	High	Med	Low	High / Med / Low	High	Linked to income
	Meat Demand	Low	Med	High	Med / Med / Med	High	Table 6
	Productivity Growth	High	Med	Low	High / Med / Low	High	Table 6
	Trade	Global	Global	Global	Regional / Regional / Local	Global	Table 6
	SPA* Policy	Afforestation	Limited afforestation	No land policy	Afforestation / Limited afforestation / No land policy	Afforestation	Table 7
Pollutant Emissions	Emissions Factors	Low	Med	High	High / High / High	Low	Emissions
Marker	Model	IMAGE	MESSAGE-GLOBIOM	AIM	GCAM	REMIND-MAGPIE
	Reference	Van Vuuren et al. (2016)	Riahi et al. (2016)	Fujimori et al. (2016)	Calvin et al. (2016)	Kriegler et al. (2016)
Challenges	Challenges to Mitigation	Low	Medium	High	Low	High
	Challenges to Adaptation	Low	Medium	High	High	Low

Electricity Assumptions

Table 3 maps technology & SSP to either core, advanced (less expensive), or low (more expensive). Costs are assumed to be equal across regions; however, resource bases for renewable energy technologies differ across regions, leading to differences in deployment in the future. Inputs are specified in the Advanced Technology and Low Technology input files.

Table 3: Capital Costs for Electric Power Plants across SSPs

Plant Type	SSP1	SSP2	SSP3	SSP4	SSP5
Nuclear	LOW	CORE	LOW	ADV	CORE
Geothermal	ADV	CORE	LOW	ADV	CORE
PV	ADV	CORE	LOW	ADV	CORE
PV with storage	ADV	CORE	LOW	ADV	CORE
Rooftop PV	ADV	CORE	LOW	ADV	CORE
CSP	ADV	CORE	LOW	ADV	CORE
CSP with storage	ADV	CORE	LOW	ADV	CORE
Wind	ADV	CORE	LOW	ADV	CORE
Wind with storage	ADV	CORE	LOW	ADV	CORE

Fossil Fuel Assumptions

Table 4 details assumptions made for fossil fuel supply. Similar to renewables, we do not vary the resource bases across SSPs, but instead vary cost of extraction and use. In particular, we adjust the technical change coefficient on future extraction costs to reflect differences in technological conditions across SSPs. To reflect differences in social acceptance, we use a cost adder that adjusts the cost to downstream users. The cost adder for 2100 is shown in Table 3; adders are linearly interpolated from $0 in the final historical year to the 2100 value. The 2100 value is defined in relation to the carbon content of the resource; thus, a “medium” social acceptance will result in a different cost adder for different fuels. Input values are specified in the Technical Change and Environmental Cost files.

Table 4a: Fossil Fuel Assumptions across SSPs: Technical Change on Extraction Cost (% per year)

Fuel	SSP1	SSP2	SSP3	SSP4	SSP5
Coal	0.5%	0.5%	1%	0.5%	2%
Gas	0.5%	0.5%	0.5%	1%	2%
Conventional Oil	0.5%	0.5%	0.5%	1%	2%
Unconventional Oil	0%	0.5%	0.5%	2%

Table 4b: Fossil Fuel Assumptions across SSPs: Cost Adder in 2100 ($/GJ)

Fuel	SSP1	SSP2	SSP3	SSP4	SSP5
Coal	$1.37	$0.27	$0	$0.27	$0
Gas	$0.14	$0.14	$0.14	$0.71	$0
Conventional Oil	$0.20	$0.20	$0.20	$0.98	$0
Unconventional Oil	$0.21	$0.21	$0.21	$1.06	$0

Building Energy Demand Assumptions

Table 5 includes information on implementation of the SSPs for building energy demand. In particular we adjust the rate at which traditional bioenergy is phased out across SSPs. This is done in GCAM by adjusting the fuel preference elasticity, a parameter that links preferences (i.e., the share weight of the logit function) to per capita income. More negative fuel preference elasticities result in faster phase outs. This parameter is used to capture differences in intraregional inequality between SSP3 and SSP4; namely, we assume a faster phase out for the same income level in SSP3 than in SSP4. Fuel preference elasticities are specified in the following input files: SSP1 and SSP5, SSP3, and SSP4.

Table 5: Building energy demand Assumptions

	SSP1	SSP2	SSP3	SSP4	SSP5
Fuel Preference Elasticity for Traditional Bioenergy	-2.5	-2	-1	-0.75	-2.5

Agriculture and Land Use Assumptions

Table 6 includes information on implementation of the SSPs for the agriculture and land-use sector. For agricultural productivity growth, we relate the parameters used in each SSP to the GCAM default assumptions. These assumptions are based on FAO projections through 2050, then held constant at 0.25%/yr post-2050. The projections in the near-term vary by crop, region, management practice (irrigated vs. rainfed), and year. As a result, there are literally thousands of parameters in a single scenario. For ease of presentation, we describe our assumptions for the individual SSPs with respect to the GCAM default. For food demand, all SSPs assume that growth in food demand is linked to income, using historical income-calorie relationships. To capture differences in food preferences and food waste, we adjust meat consumption and preferences in different SSPs. Finally, trade restrictions on agriculture and bioenergy are imposed in the SSP3 and SSP4. Inputs for each are found in the following input files / data system processing files: Meat Demand, Agricultural Productivity Growth, and Trade.

Table 6: Agriculture and Land Use Assumptions across SSPs

Variable	SSP1	SSP2	SSP3	SSP4 (High / Med / Low Income )	SSP5
Agricultural Productivity Growth	50% above GCAM default	GCAM default	50% below GCAM default	50% above GCAM default / GCAM default / 50% below GCAM default	50% above GCAM default
Meat Demand	Demand linked to income and price. Shift in preferences among meat products away from beef, sheep, and goats.	Demand linked to income and price.	Demand linked to income and price.	Demand linked to income and price.	Demand linked to income and price.

Climate Policy

To move from baseline scenarios to the RCP replications, we implemented the Shared Policy Assumptions (Kriegler et al., 2014) in GCAM. These assumptions are described in Table 7. For land policy, the relevant policy is phased in over several decades once the globally harmonized carbon price is imposed. For CO₂ prices prior to global cooperation, we impose the carbon price required to reach a region’s Copenhagen pledge in an SSP2 world in all SSPs. That is, the same carbon price in 2020 is used regardless of SSP. Note that these policy assumptions are not unique and alternative assumptions could be derived that are also consistent with the SSP storylines. Note that the land carbon price described in the table is absent any transaction costs; we further reduce carbon prices on land by 50% to capture these effects.

Table 7a: Policy Assumptions across SSPs: First model time period with global cooperation (i.e., harmonized global carbon price)

Income Group	SSP1	SSP2	SSP3	SSP4	SSP5
High Income	2025	2040	2040	2025	2040
Medium Income	2025	2040	2040	2025	2040
Low Income	2025	2040	2050	2025	2040

Table 7b: Policy Assumptions across SSPs: Land Policy

Income Group	SSP1	SSP2	SSP3	SSP4	SSP5
High Income	Carbon Price on Land is Equal to Energy Carbon Price (i.e., UCT in Wise et al., 2009)	Carbon Price on Land is Equal to 50% of Energy Carbon Price	No Land Policy (i.e., FFICT in Wise et al., 2009)	Carbon Price on Land is Equal to Energy Carbon Price (i.e., UCT in Wise et al., 2009)	Carbon Price on Land is Equal to Energy Carbon Price (i.e., UCT in Wise et al., 2009)
Medium Income	Carbon Price on Land is Equal to Energy Carbon Price (i.e., UCT in Wise et al., 2009)	Carbon Price on Land is Equal to 50% of Energy Carbon Price	No Land Policy (i.e., FFICT in Wise et al., 2009)	Carbon Price on Land is Equal to 50% of Energy Carbon Price	Carbon Price on Land is Equal to Energy Carbon Price (i.e., UCT in Wise et al., 2009)
Low Income	Carbon Price on Land is Equal to Energy Carbon Price (i.e., UCT in Wise et al., 2009)	Carbon Price on Land is Equal to 50% of Energy Carbon Price	No Land Policy (i.e., FFICT in Wise et al., 2009)	No Land Policy (i.e., FFICT in Wise et al., 2009)	Carbon Price on Land is Equal to Energy Carbon Price (i.e., UCT in Wise et al., 2009)

Differences between Official SSPs and SSPs using the current model version

The official SSPs (Calvin et al., 2017) were branched from GCAM4.0. There are several changes between versions that will have implications for SSP-related results. First, in the original SSPs (Calvin et al., 2017), electricity power plant costs were based off of the values used in GCAM4.0. We defined advanced and low as deviations from the core model. GCAM4.2, however, updated the power plant capital costs in the core, and established advanced and low technology costs. These values will be used going forward in the SSPs. Second, GCAM5.1 includes different land regions and multiple land technologies (see below). These are likely to have effects on all land-related results. Third, GCAM4.4 introduced a net negative emissions constraint that wasn’t present in the official SSPs. This will affect bioenergy use and net negative CO₂ emissions, particularly in 2.6 W/m² scenarios.

Data from the official SSPs is available at SSP database.

References

Calvin, K., B. Bond-Lamberty, L. Clarke, J. Edmonds, J. Eom, C. Hartin, S. Kim, P. Kyle, R. Link, R. Moss, H. McJeon, P. Patel, S. Smith, S. Waldhoff and M. Wise (2017). “The SSP4: A world of deepening inequality.” Global Environmental Change 42: 284-296.

Dellink, R., et al., Long-term economic growth projections in the Shared Socioeconomic Pathways. Global Environ. Change (2016), http://dx.doi.org/10.1016/j.gloenvcha.2015.06.004

Kriegler, E., Edmonds, J., Hallegatte, S., Ebi, K. L., Kram, T., Riahi, K., Winkler, H. & Van Vuuren, D. P. 2014. A new scenario framework for climate change research: The concept of shared climate policy assumptions. Climatic Change, 122, 401-414.

Lutz, KC, S., W., The human core of the shared socioeconomic pathways: Population scenarios by age, sex and level of education for all countries to 2100. Global Environ. Change (2016), http://dx.doi.org/10.1016/j.gloenvcha.2014.06.004

Moss, R. H., J. A. Edmonds, K. A. Hibbard, M. R. Manning, S. K. Rose, D. P. van Vuuren, T. R. Carter, S. Emori, M. Kainuma, T. Kram, G. A. Meehl, J. F. B. Mitchell, N. Nakicenovic, K. Riahi, S. J. Smith, R. J. Stouffer, A. M. Thomson, J. P. Weyant and T. J. Wilbanks (2010). “The next generation of scenarios for climate change research and assessment.” Nature 463(7282): 747-756.

O’Neill, Brian C., Elmar Kriegler, Kristie L. Ebi, Eric Kemp-Benedict, Keywan Riahi, Dale S. Rothman, Bas J. van Ruijven et al. “The roads ahead: Narratives for shared socioeconomic pathways describing world futures in the 21st century.” Global Environmental Change (2016). doi:10.1016/j.gloenvcha.2015.01.00

Riahi, K., D. P. van Vuuren, E. Kriegler, J. Edmonds, B. C. O’Neill, S. Fujimori, N. Bauer, K. Calvin, R. Dellink, O. Fricko, W. Lutz, A. Popp, J. C. Cuaresma, S. Kc, M. Leimbach, L. Jiang, T. Kram, S. Rao, J. Emmerling, K. Ebi, T. Hasegawa, P. Havlik, F. Humpenöder, L. A. Da Silva, S. Smith, E. Stehfest, V. Bosetti, J. Eom, D. Gernaat, T. Masui, J. Rogelj, J. Strefler, L. Drouet, V. Krey, G. Luderer, M. Harmsen, K. Takahashi, L. Baumstark, J. C. Doelman, M. Kainuma, Z. Klimont, G. Marangoni, H. Lotze-Campen, M. Obersteiner, A. Tabeau and M. Tavoni (2017). “The Shared Socioeconomic Pathways and their energy, land use, and greenhouse gas emissions implications: An overview.” Global Environmental Change 42: 153-168.

van Vuuren, D. P., E. Kriegler, B. C. O’Neill, K. L. Ebi, K. Riahi, T. R. Carter, J. Edmonds, S. Hallegatte, T. Kram, R. Mathur and H. Winkler (2014). “A new scenario framework for Climate Change Research: scenario matrix architecture.” Climatic Change 122(3): 373-386.