New: Pivotal Year—WRI’s 2015 Annual Report

You are here

INSIDER: Why Are INDC Studies Reaching Different Temperature Estimates?

On Friday, the UNEP Emissions Gap Report joined a series of studies released over the past few weeks assessing how much countries’ recent climate change announcements, or intended nationally determined contributions (INDCs), contribute to combating warming. Collectively, the studies make it clear that the INDCs make a substantial contribution to bending the global emissions trajectory below our current path. However, the studies also show that without additional action, the INDCs are insufficient to limit warming to below 2°C and avoid some of the worst climate impacts. The details of the Paris Agreement are, therefore, very important to help achieve an additional bending of the emissions trajectory before 2020, to support the implementation of the INDCs and to ensure greater ambition after 2030.

While all of the studies support these general conclusions, their numbers differ both in terms of projected temperature increase relative to pre-industrial levels, as well as emissions levels in 2025 and 2030. Below we discuss the results of various studies assessing the INDCs and highlight differences among their underlying assumptions, which can help explain why some details of their findings diverge. The studies include those by Climate Action Tracker (CAT), Australian-German Climate and Energy College (CEC), Climate Interactive, Danish Energy Agency (DEA), European Commission Joint Research Centre (EC-JRC), the International Energy Agency (IEA), London School of Economics (LSE), Massachusetts Institute of Technology (MIT), MILES Project Consortium (MILES), PBL Netherlands Environmental Assessment Agency, the UNFCCC, and the UNEP Emissions Gap Report, which is itself an in-depth assessment of many of these studies. Some of these studies may still be updated after this blog is published.

INDCs are driving emissions down, but not fast enough.

Most studies compare global emissions pathways resulting from the INDCs to one or more alternative scenarios without the INDCs. These alternative scenarios include emissions under “business as usual” (typically defined as no new climate policy from 2010 onwards), emissions under currently adopted and implemented policies, and emissions assuming that countries’ 2020 pledges are met. The studies find that the INDCs reduce future emissions relative to these scenarios.

INDCs put us on track for a world that is 2.7-3.7 degrees C warmer (median chance), depending on modeling assumptions.

The studies also find, however, that the INDCs do not reduce emissions sufficiently to limit warming to below 2 degrees C. Some of the studies model global temperature change under the INDCs out to 2100, and find that it will be higher than 2 degrees C. Other studies do not model temperature directly, but they still draw conclusions about temperature. They do this by comparing emissions in 2025 and/or 2030 under the INDCs to the levels that would be consistent with limiting warming to below 2C at the least possible cost. Either way, the result is the same - the INDCs are not sufficient.

The subset of studies that assess temperature increases suggest that with the INDCs, we will witness 2.7-3.7 degrees C (median chance) of warming compared with pre-industrial levels. This is an improvement over business-as-usual trends, which would lead to 4-5 degrees C of warming, but falls short of the goal to limit warming to below 2 degrees C. Since temperature impacts are calculated out to 2100, the studies’ findings depend significantly on assumptions about what happens to emissions after the target date specified in the INDC  -- 2030 for most countries, and 2025 for the United States.

Scenarios showing higher temperature increases by 2100, such as Climate Interactive INDC Strict, assume no continued progress after the INDCs are achieved. The scenarios showing lower temperature increases, such as the Climate Action Tracker, assume that mitigation effort of 2020-2030 continues throughout the century. There are also additional scenarios that have been studied by some analysts (e.g. the MILES study’s “Bridge Scenario”) that examine what additional targets and policies would have to be adapted to limit warming to 2 degrees C.



Study

Scenarios assessed

Climate Action Tracker

  • The 50% probability scenario gives a median chance that the INDCs will lead to 2.7 degrees C.
  • The >66% probability scenario results in a “likely” chance of the INDCs leading to 3 degrees C.

Climate Interactive

  • INDC Strict – No change after national contribution pledge period

EC-JRC

  • INDC-unconditional: All INDCs expressed unconditionally are implemented; countries where the reference already lead to emissions at or lower than their INDCs, as well as countries with no INDCs or conditional-only INDCs, do not implement additional policies. No commitment was assumed for low-income African countries. Beyond 2030, regional carbon prices increase, including for countries that previously had no climate policies, and progressively converge at a speed that depends on their per capita income; on average, the world GHG intensity over 2030-2050 decreases at the same rate as for 2020-2030.
  • INDC-conditional: Similar to INDC-unconditional, but all INDCs are implemented, including all conditional contributions.
  • The emissions in the two INDC scenarios result in a global temperature increase of around 3 °C. The INDC scenarios fall in the middle of the range defined by Representative Concentration Pathway RCP4.5, corresponding to a likelihood of "staying below 3 °C over the 21st century" of "likely" and "more likely than not.”

IEA

  • The IEA has analyzed the INDCs, together with the latest available global energy data (including recent official revisions to China’s energy balance). Beyond the INDCs, the analysis takes into account all domestic energy sector policies that are currently in place or under discussion across all countries. The analysis is based upon the full implementation of unconditional INDC pledges. Some countries have also indicated that they might agree a more ambitious INDC under certain conditions, but such additional pledges are not included here. To assess the impact on global average temperature increase, IEA used MAGICC with an emissions pathway in between the representative concentration pathways (RCP) 4.5 and 6 from the IPCC’s Fifth Assessment Report. This was judged as the long-term emissions trajectory most closely aligned with this INDC analysis.

MILES

  • While a 2°C- bridge scenario allowing for a more continuous transition from the INDCs to 2°C, and a global INDC--2°C which implements INDCs until 2030 and then shifts to a 2°C scenario from 2030, were modeled, included here is the INDC-extended scenario. In this scenario, the INDCs were implemented on the level of the 11 REMIND regions, and extended beyond 2030 by an extrapolation of the regional carbon prices that emerged under the INDCs, as well as technology targets for selected regions. We excluded the other scenarios because they implement additional targets and policies to limit warming to 2°C.
  • The 3 degrees C result hinges on post-2030 policy assumptions of the scenario, e.g. see the regional carbon prices as illustrated in Figure A.1 in their analysis.

MIT

  • Low, central and high estimates for, respectively, low, median, and high climate sensitivity scenarios. The analysts choose three values of climate sensitivity (CS) that correspond to the 5th percentile (CS=2.0°C), median (CS=2.5°C), and 95th percentile (CS=4.5°C) of the probability density function that were jointly estimated with the ocean heat uptake rate.

All temperature findings are associated with some degree of uncertainty, so analysts use probabilities to frame their results. Two common framings are 50% likelihood, which provides a “toss of the coin” chance that warming will stay within the given temperature, and >66% likelihood, which provides a “likely” chance that warming will stay within the given temperature. The Climate Action Tracker finds a 50% chance that warming will stay within 2.7C, and a >66% chance that it will stay within 3C. Climate Interactive Strict finds that under its INDC Strict scenario there is a 50% chance that warming will be limited to 3.5C. MIT presents three scenarios – the low scenario corresponds to the 5th percentile; the central scenario corresponds to the median, and the high corresponds to the 95th percentile of the probability density function. Given that policy makers should be interested in a higher probability of achieving desired temperature outcomes, more studies should examine what a higher probability of the resulting temperatures from the INDCs are. A higher probability of limiting warming to various temperatures gives much greater confidence that the INDCs will be successful in limiting warming to that temperature.

2025 and 2030 emissions levels matter for warming.

Emissions levels in 2025 and 2030 have significant consequences for our ability to limit warming to 2 degrees C.  The higher emissions are in the near term, the greater the required emissions reductions in later decades for limiting warming. Steep rates of emissions reductions are far costlier than more gradual rates of decline. They also risk of failing to achieve the 2 degrees C target, and rely more on carbon dioxide removal technologies (e.g. bioenergy and carbon capture and storage), which have yet to be proven at scale. 

The IPCC Fifth Assessment Report finds that if emissions levels are above 55 Gt CO2e in 2030, they require 6 percent per year rates of emissions decline between 2030 and 2050 (compared with 3 percent/year in cost-effective scenarios). A 6 percent rate of emissions reductions is unprecedented—emissions reduction rates during the collapse of the Soviet Union led to declines of 2-4 percent annually—and it will be exceedingly difficult to overcome the lock-in of carbon-intensive technologies, while at the same time rapidly scaling up zero and low-carbon energy sources.

All studies included in our analysis find that emissions levels in 2025 and 2030 are higher than those consistent with a likely chance of limiting warming to 2 degrees C. The emissions levels in the studies range from 51.1-57.2 Gt CO2e in 2025 and 52-61.1 Gt CO2e in 2030. For contrast, the UNEP Emissions Gap Report finds that for a least-cost emissions pathway consistent with a likely chance of limiting warming to 2 degrees C, emissions are 48 Gt CO2e in 2025 and 42 Gt CO2e in 2030. Below is a chart showing median values for various studies’ assessments of anticipated emissions levels in 2025 and 2030, given the INDCs. The chart does not include those scenarios assuming mitigation targets and policies that went above and beyond the INDCs, nor does it include the IEA, which reports energy and process-related greenhouse gas emissions.

Some of the studies examine multiple scenarios – for example, one scenario might include those pledges that are conditional (for example, on international finance or other support), while another might include only unconditional pledges. Below we describe the differences among multiple scenarios.


Study

Scenarios assessed

Australian-German CEC

  • The conditional INDCs are included in “CEC high”, not in the “CEC low” scenario. The difference between the estimates is also due to which countries present unconditional ranges.

EC-JRC

  • INDC-unconditional: All INDCs expressed unconditionally are implemented; countries where the reference already lead to emissions at or lower than their INDCs, as well as countries with no INDCs or conditional-only INDCs, do not implement additional policies. No commitment was assumed for low-income African countries. Beyond 2030, regional carbon prices increase, including for countries that previously had no climate policies, and progressively converge at a speed that depends on their per capita income; on average, the world GHG intensity over 2030-2050 decreases at the same rate as for 2020-2030.
  • INDC-conditional: Similar to INDC-unconditional, but all INDCs are implemented, including all conditional contributions.

LSE

  • INDC Scenario 1 shows the level of emissions consistent with meeting the INDC targets if independent GDP growth rate projections from the IMF and OECD are used for the intensity target calculations (see footnote 10 for more information).
  • INDC Scenario 2 shows the same result if official national GDP growth rate projections are used for the intensity target calculations for China and India.
  • INDC Scenario 3 adopts the ambitious end of target ranges and all conditional targets.

MILES

  • While a 2°C- bridge scenario allowing for a more continuous transition from the INDCs to 2°C, and a global INDC--2°C which implements INDCs until 2030 and then shifts to a 2°C scenario from 2030, were modeled, included here is the INDC-extended scenario. In this scenario, the INDCs were implemented on the level of the 11 REMIND regions, and extended beyond 2030 by an extrapolation of the regional carbon prices that emerged under the INDCs, as well as technology targets for selected regions. We excluded the other scenarios because they implement additional targets and policies to limit warming to 2°C.

PBL

  • Unconditional INDC: Includes INDCs that are unconditional.
  • Conditional INDC: Further includes INDCs that are conditional.

Notes: The EC-JRC study excludes sinks.



Why are these studies reaching different estimates?

There are a number of factors that can explain why studies estimate different temperature outcomes and emissions levels. For some countries, estimating future emissions is straightforward, and results in relatively little difference across studies (e.g. those countries that pledge to reduce economy-wide emissions relative to a past year’s emissions). For others, analysts have to make more assumptions (e.g. countries that pledge to reduce emissions per unit of GDP, without specifying expected future GDP growth). The emissions covered by INDCs are then aggregated with projected future emissions from countries, sectors and gases not covered by INDCs. The latter are taken from projections of what future emissions will be under “business as usual” or under current policies.

Three factors are largely responsible for these differences:

  1. Which INDCs the analysis covers
  2. Depending on the timing of the analysis, the number of INDCs analyzed differs. For example, of the studies we looked at, the cut-off dates for including INDCs ranges from mid-August to the end of October 2015.

    ---------

    Cut-Off Date For Analysis And Countries Included, In Order Of Earliest To Latest

    Study

    Cut-off date

    MIT

    Mid-August 2015

    PBL

    8/31/2015

    CAT

    10/1/2015

    Australian-German CEC

    10/30/2015

    Climate Interactive

    10/20/2015

    UNFCCC

    10/1/2015

    MILES

    10/2/2015

    DEA

    10/1/2015

    EC-JRC

    10/13/2015

    IEA

    Mid-October 2015

    LSE

    10/23/2015

    ---------

    In addition, a number of countries attach conditions (such as international financing or other forms of support) to all or part of their INDCs. Some scenarios include unconditional pledges only, whereas others include both conditional and unconditional pledges. Moreover, some INDCs are not clear about the extent to which they are conditional; in these cases, studies may categorize the same pledge differently with respect to its conditionality.

    ---------

    Treatment Of Conditionality Of INDCs Across Studies

    Study

     

    Australian-German CEC

    The conditional INDCs are included in “CEC high”, not in the “CEC low” scenario.

    CAT

    Temperature findings are based on unconditional pledges only

    Climate Interactive

    Unconditional INDCs only

    DEA

    Central estimate based on a range of scenarios, each of which includes a variant with and without the conditional pledges.

    EC-JRC

    INDC-unconditional: Unconditional only

    INDC-conditional: Conditional and unconditional

    IEA

    The analysis is based upon the full implementation of unconditional INDC pledges. Some countries have also indicated that they might agree a more ambitious INDC under certain conditions, but such additional pledges are not included here.

    LSE

    Scenarios 1 and 2: Exclude conditional targets for countries that explicitly have both conditional and unconditional targets.

    Scenario 3: Include all conditional targets.

    MILES

    Country-level analysis of conditional INDCs from PBL

    MIT

    Unconditional only

    PBL

    Conditional and unconditional INDCs are modeled in separate scenarios

    UNFCCC

    Both conditional and unconditional included within a single range

    ---------
  3. Assumptions about the INDCs
  4. Although the INDCs are more transparent than the Cancun pledges, the variation in studies’ results is in part due to the fact that some countries have not identified an expected emissions level in the future. This is particularly the case for targets that are framed as intensity targets or baseline scenario targets, where projected GDP or baseline scenario emissions are not specified, or targets that set a year to peak emissions and don't specify the peak emissions level.

    For example, how China's emissions trajectory is treated will have a significant impact on global emissions and resulting temperature increase, given that it is responsible for 22% of global emissions. With the exact timing and level of emissions peak is not specified, studies are forced to make assumptions about China's future emissions levels. In addition, some studies examine China's intensity target only, while others also examine additional policies and technology transformation, which can lead to additional reductions. Furthermore, given that the Chinese INDC peak year target only covers carbon dioxide, analysts have to make assumptions about the growth rate of non-CO2 gases.


Assumptions About China’s Emissions Under The INDC

Study

Treatment of China

Australian German-CEC

2025: 14.1 Gt CO2e (low) 13.0 Gt CO2e (high)

2030: 14.7 Gt CO2e (low) 13.3 Gt CO2e (high)

Assumes China’s intensity targets are taken with GDP projections as used in Fu Sha et al.

CAT

2030 emissions excluding land use, land-use change, and forestry (LULUCF): 13.6 GtCO2e

CO2 emissions peak between 2025 and 2030.

Setting aside the carbon intensity target, China’s INDC’s actions and non-fossil energy target lead to GHG emission levels of around 13.6 GtCO2e in 2030. The intensity target, if dominating other elements of the INDC, national policies and actions, would lead to much higher 2030 emission levels of 15- 16.9 GtCO2e.

Climate Interactive

CO2 emissions flat-line at 15% below BAU in 2030 (peaking at 14.4Gt/year), other GHGs grow at BAU

LSE

INDC 1: 12.8-14.3 GtCO2e

INDC 2: 14.5-16.3 GtCO2e

Emissions improvement and GDP projections (approximately 6.2% annual growth 2015–2020, 4% 2020–2030) based on projections from the IMF (2015) and OECD (2014).  CO2: 2005 baseline for all CO2 is 5.9 GtCO2e from China’s 2nd National Communications report. Non-CO2: hypothetical Pre-INDC Reference Scenario projection used only for non-CO2 gases: 2.7 GtCO2e, and LULUCF of -284 MtCO2e. Target 60–65% improvement in emissions intensity from 2005 gives 12.8–14.3 GtCO2e of emissions in 2030 (INDC Scenario 1). With alternative GDP assumptions (approximately 7% annual growth 2015– 2020, 5.3% 2020–2030) taken from analysis by the Chinese National Climate Strategy Centre (NCSC). Target emissions in 2030 are calculated as 14.5–16.3 GtCO2e (INDC Scenario 2).

MIT

2025: 13.3 Gt CO2e  (total GHG net of land use); -3 MT CO2 (land use change CO2)

2030: 13.1 Gt CO2e (total GHG net of land use); 221 MT CO2 (land use change CO2)

China’s fossil CO2 emissions are capped at 9.4 Gt CO2 after 2030. Coal consumption (primary energy) peaks in 2025 at 66.3 EJ.

PBL

 2025: 13.9 GtCO2e

 2030: 14.0 GtCO2e

UNFCCC

China’s emissions not specified, but following methodological points are relevant:

for intensity targets, the estimated level of emissions in the target year was calculated by, firstly, subtracting from the intensity in the reference year the percentage specified by the Party for that target year and, secondly, by multiplying the resulting intensity by the expected level of GDP in the target year, as communicated by the Parties, if available; for emission-peaking targets, historical emission growth rates were projected linearly towards zero in the year of peaking to obtain an estimate of maximum emissions; for other types of INDCs, including policies and measures, this report does not contain a quantification of their effect unless official estimates for emissions in 2025 and 2030 were provided by the Party in question. For Parties that used a combination of different type of targets, expected levels of emissions in 2025 and 2030 were estimated individually. The target that resulted in the lowest emission levels was used in the calculation of the aggregate emissions (Whenever necessary, multiple initial growth rates were used. The secretariat ensured that the estimates were consistent with national expert assessments.)


In addition, studies rely upon different data sources both for historical emissions as well as projected emissions (which is particularly relevant to the analysis of uncovered sectors and gases, as well as countries that have not submitted an INDC). Official data submitted by countries can vary from international data sources that are harmonized across countries. For instance, we found that 2010 emissions estimates used by various studies ranged from 47-49.5 Gt CO2e. In addition, studies use projections of emissions to estimate future emissions for those countries, sectors and gases not covered by an INDC. The difference between relying on “business as usual” projections versus current policy projections can be significant. Furthermore, some studies use IPCC Second Assessment Report global warming potential (GWP) values while others use Fourth Assessment Report, which can lead to differences in emissions estimates for non-CO2 gases.


2010 Emissions Estimates Across Studies

Study

2010 emissions (Gt CO2e)

Australian-German CEC

48.4

CAT

47

Climate Interactive

50.7

DEA

48.0

EC-JRC

48.2 (excluding sinks)

MILES

48.8

MIT

45.7 (net of land use); 2.6 land use CO2

LSE

49.8

PBL

49.5

UNFCCC

48.1


Also, the way in which emissions reductions and enhanced sinks are accounted for can have an impact on future emissions levels. This is particularly the case for LULUCF accounting. Projections of future emissions and removals in the sector are extremely uncertain, and most countries did not specify their assumed accounting approaches or data sources. The DEA study, for example, finds that differences in accounting rules for LULUCF can increase the emissions gap in 2030 by around 0.8 to 3.4 GtCO2e.

Assumptions about what happens after the INDCs

One of the most significant factors is likely to be what the studies assume after the target year of the national climate plans, which determines the countries’ post-2025 or post-2030 emissions trajectory. For instance, do actions continue or end after the INDC is completed? Since most INDCs do not specify action after 2030, different scenarios reflect a broader range of assumptions, resulting in divergence in conclusions regarding post-2030 global emissions. The Climate Action Tracker assumes continued emissions reductions in future decades at a level of effort no more, or less, ambitious than that implied by the INDCs. Climate Interactive, on the other hand, assesses multiple scenarios, one in which there is no post-2030 action and several in which there is continued emissions reductions in certain countries to varying extents. The MIT study assumes the proposed cuts from the INDCs are extended through 2100, but not deepened further.


Assumptions About Post-INDC Emissions Trajectory

Study

Post-INDC climate action

CAT

Reflects continued emissions reductions at a level of effort equal to that implied by the INDCs. The relative position of the INDC emissions pathway within a distribution of emissions pathways generated from the AR5 database is used as a proxy for the level of effort. This position (the quantile) is kept constant after 2030 to determine the post-INDC emissions trajectory. The algorithm is applied to each of the 5 RCP regions individually. If countries have 2050 emission targets, these are included in the analysis (e.g. USA, EU, and Japan). The extension of the emissions trajectory is then carried out starting from 2050 in the same way as for the 2030 case.

Climate Interactive

For absolute reductions, emissions stay constant after INDC period, unless INDC states otherwise. For reductions relative to a reference scenario, emissions stay at the target level below the reference scenario after the INDC period. For peak year targets, emissions remain constant at the peak level. For intensity targets, emissions intensity remains at the targeted level after the INDC period.

EC-JRC

INDC Low: Emissions peak in 2035

INDC High: Emissions peak in 2030

IEA

If climate ambition is not raised progressively, it is estimated that the path set by the INDCs would be consistent with an average global temperature increase of around 2.7 degrees Celsius (°C) by 2100, falling short of limiting the increase to no more than 2 °C. To assess the impact on global average temperature increase, we used MAGICC with an emissions pathway in between the representative concentration pathways (RCP) 4.5 and 6 from the IPCC’s Fifth Assessment Report. This was judged as the long-term emissions trajectory most closely aligned with this INDC analysis.

MILES

INDC-extended: INDCs implemented until 2030, with extrapolation of energy sector and carbon pricing policies afterwards.

Global INDC-2°C: INDCs implemented until 2030, and then unanticipated shift to 2°C policies after 2030

2°C- bridge scenario: Policies and targets are strengthened by 2020 for 2030 and beyond, and investors respond early to this strengthened policy commitment. After 2020, full anticipation of 2°C-compatible carbon pricing to be in place after 2030, with investors responding immediately to this strengthened policy commitment.

MIT

Assumes the proposed cuts are extended through 2100 but not deepened further; where

INDCs specify actions through 2030, assumes these levels of commitment remain in place throughout the study. For other regions, represents Copenhagen–Cancun commitments throughout the study.

(Studies not listed do not extend past 2030)


In addition, other factors could make a difference, such as assumed implementation of the pledges, variations among the models themselves, treatment of accounting rules, and more.

The Potential of the Paris Agreement

The Paris Agreement can help bend the curve further before 2030 and ensure greater ambition after 2030 by including clear long-term and short-term signals, increasing transparency of INDCs and future cycles of commitments, and advancing accounting rules governed by strong principles. Additionally it can increase the probability of full implementation through strong provisions on capacity building, finance and technology transfer. In a little less than a month’s time, we have the opportunity to build upon the momentum unleashed by the INDCs and start to close the emissions gap to have a fighting chance of keeping global average temperature below 2 degrees C.

INSIDER, cop21

Comments

Very interesting and some of useful dadas! Arigatouuu ! I share!
Badman Nishioka/Japan/rainforest action group/HUTAN Group/

Add new comment

Stay Connected