Why a scoreboard, and what are you keeping score of?
The question is: are current government proposals ambitious enough to limit global warming to well below 2°C (3.6°F), with an effort toward 1.5°C?
In 2010 at the United Nations Framework Convention on Climate Change (UNFCCC) COP16 negotiations in Cancun, nations agreed to limit greenhouse gas (GHG) emissions to prevent temperature rise to no more than 2°C (3.6°F) with some countries calling for the goal to be set lower at 1.5°C. Parties to the UNFCCC are poised to meet again for the 21st Conference of Parties (COP21) in Paris, France in December, 2015 and are submitting pledges to reach this goal. The Climate Scoreboard shows the progress nations have made.
Calculating the impact of the various proposals is dependent upon many variables including population growth, economic activity and the decisions of governments around the world and how they affect the complex dynamics of the Earth’s climate system. Even with a clear understanding of those factors, we must use computer modeling to explore the long-term impacts of any given emissions reduction proposal on greenhouse gas concentrations and temperature.
We created the C-ROADS simulator to address this challenge. C-ROADS can add up emissions reductions (for fossil fuel emissions, other GHG emissions, and land-use emissions) from proposals into a single, global emissions trajectory. The model then accounts for the cycling of the gases through the biosphere and atmosphere and calculates their impacts on the Earth’s mean temperature over time. Our team also uses C-ROADS as the calculation engine behind the Climate Scoreboard. However, C-ROADS is used in many contexts – as a tool for teaching and for interactive exercises and as a decision support tool for policy makers (including some UNFCCC negotiators).
How do you calculate the results?
C-ROADS allows inputs of emissions reductions targets for 12 countries (U.S., Japan, Russia, Canada, Australia, South Korea, China, India, Brazil, Mexico, South Africa, Indonesia) and three blocs of countries (EU, “other developed” and “other developing”). For the 12 countries with individual controls, we adjust that country’s modeled emissions from a reference scenario (based on the IPCCs RCP 8.5 scenario).
The emissions from EU countries are changed as a bloc. Pledges of countries not technically included in the EU (e.g., Andorra) and of countries within the EU that submitted their own NDC are properly scaled to their share of 2013 emissions within the EU bloc.
NDCs accounted for in “other developed” or “other developing” were calculated in accordance with that country’s percentage of emissions from the entire group. In other words, each nation’s proposals are properly scaled to their actual share of 2013 emissions within their bloc.
What assumptions and approximations do you make in this analysis?
Any effort like ours, which is trying to reduce a complicated real world system into a few simple indicators, must make approximations and assumptions. In doing so, we do our best to be aware of and make explicit as many of those assumptions as we can.
Assumptions about proposals:
We include only Nationally Determined Contributions (NDCs) submitted to the UN Framework Convention on Climate Change (UNFCCC) that are unconditional. We assume that proposals will be fully implemented, in accordance with the dates specified in the NDC.
There are four types of pledges for which we made post-pledge-period (2025-2030) assumptions for the “NDCs Strict” case:
- Absolute reductions. For countries that make absolute emissions reductions, we assume that emissions stay constant after 2025-2030 (the end of the pledge period), unless an NDC states otherwise. For example, Canada has offered to reduce their GHG emissions 30% below 2005 levels by 2030. After that, given there is no post-2030 pledge, we assume that the emissions do not increase nor continue to decrease, but rather stay constant at their 2030 levels.
- Relative to a reference scenario. For countries that are reducing emissions from a reference scenario or business as usual (BAU) level, we assume they stay at the target level below the reference scenario after the pledge period. For example, Kenya has pledged to reduce their emissions 30% below their BAU trajectory by 2030. After 2030, we assumed that the emissions would continue to stay 30% below their BAU Scenario. Since the BAU Scenario is increasing, emissions increase at the same rate as the Reference Scenario.
- Peaking. Emissions grow to a target level. For example, China has said they will peak their CO2 emissions by 2030. We assumed that the rate of growth of emissions would slow down starting in 2015, approaching a flat trajectory gradually. After the peaking date, 2030 in this case, emissions would stay level. (Note — other greenhouse gas emissions continue to increase given they are not covered by China’s pledge).
- Emissions intensity reduction. This is would be a change in emissions relative to the GDP of the country. For example, Singapore intends to reduce their emissions intensity 36% by 2030. Up to 2030, we calculate emissions by using the GDP growth. After 2030, emissions intensity remains at that target intensity unless it exceeds the BAU intensity, at which point the BAU trajectory resumes (GDP assumption is from SSP2–Shared Socioeconomic Pathways).
Some countries have made proposals for specific actions (rather than for emissions targets). If we don’t have a good estimate for the impact of those actions on emissions then we don’t include the action in our analysis.
We assume emission reductions cover all GHG emissions (CO2 from fossil fuels, CO2 from LULUCF, CH4, N2O, SF6, PFCs, and HFCs), unless specified otherwise. The model can apply emissions reductions to either all or individual GHGs; when the NDC applies only to CO2, the model assumes BAU for other GHGs. We apportion cuts across all the gasses and ensure that they add up to the appropriate reduction in CO2e. If a pledge applies only to CO2, then the other gases follow their BAU trajectories.
Assumptions inherent in the C-ROADS model:
- The C-ROADS “reference scenario” (often called “business as usual”) is based on the RCP 8.5 scenario from the Intergovernmental Panel on Climate Change’s (IPCC) Fifth Assessment Report (AR5) report.
- C-ROADS is generally set to a climate sensitivity of 3 (that is 3°C of temperature increase for a doubling of CO2) However, that assumption can be varied in the model, if desired.
- C-ROADS is a physics-based model that includes plenty of climate science. There are positive feedback loops in the real climate system that are not modeled in the current version of C-ROADS. Additionally, C-ROADS is based upon and calibrated to the results of models from the IPCC’s AR5.
- For more information on the many other assumptions in C-ROADS, please consult the C-ROADS Reference Guide.
How often are you updating the results?
We will update the results shown on the Climate Scoreboard in the lead-up to the Paris climate negotiations if Nationally Determined Contributions (NDCs) are submitted that result in changes of at least 0.1° C to the temperature in 2100.
How did you determine the extent of the blue “uncertainty range” on the Climate Scoreboard thermometer?
We analyzed the confidence ranges for the different scenarios in the IPCC’s AR4 report, ranging from B1 to A1FI and calculated the range of temperatures across the suite models relative to the mean temperature change estimate. We estimated that the 90% confidence interval (defined as “likely” by AR4), and show that range as the uncertainty in the Climate Scoreboard.
The line in the middle that we reference most prominently is the “mean” or average value.
Is your model scientifically reviewed?
The C-ROADS model was scientifically reviewed by a panel of outside, independent experts, chaired by Sir Robert Watson (former chair of the IPCC).
Who else has done similar calculations and why are they sometimes different?
There are many groups that have done similar analyses of the temperature impact of the pledges. Taryn Fransen and Kelly Levin at World Resources Institute have summarized some of the differences between these estimates in their post: Why Are INDC Studies Reaching Different Temperature Estimates?
Other resources include: