Bob Kopp

Climate and sea-level scientist and climate policy scholar at Rutgers University’s Dept. of Earth & Planetary Sciences and Rutgers Climate & Energy Institute.

12 minute read

August 27, 2026

Dr. Matthew Wielicki
Director, U.S. Global Change Research Program
Office of Science and Technology Policy

Re: RFI Response: Proposed Amendment to NCA5 (91 FR 54893)

Dear Dr. Wielicki:

I am writing to comment upon the proposed amendment to the Fifth National Climate Assessment (NCA5), subtitled “Implausible High-Emission Scenario Dependence, Gold Standard Science, and Observational Grounding.”

The proposed amendment should be withdrawn in its entirety. While it correctly identifies that very high emissions scenarios, such as RCP8.5 and SSP5-8.5, are not “expected, baseline, business-as-usual, likely, [or] central” scenarios, it does not provide a scientifically sound, adequately documented, or procedurally appropriate basis for modifying NCA5. Questions regarding the implications of recent scenario literature should instead be evaluated through the statutorily required development of the Sixth National Climate Assessment (NCA6), using the transparent, multidisciplinary, peer-reviewed assessment process established by the Global Change Research Act and following the requirements of the Information Quality Act.

1. The amendment is correct that RCP8.5 and SSP5-8.5 are not central scenarios. NCA5 already recognizes this.

The amendment is correct that RCP8.5 and SSP5-8.5 should not be interpreted as “expected, baseline, business-as-usual, likely, [or] central” 21st century emissions future. RCP8.5 was developed to represent the radiative forcing associated with the 90th percentile of reference scenarios in the published literature as of September 2007, so from its conception was not intended to represent ‘expected’, ‘likely’ or ‘central’ emissions.1 The 8.5 W/m2 2100 forcing value associated with SSP5-8.5 was chosen to provide continuity with RCP8.5. Thus, RCP8.5 and SSP5-8.5, in terms of radiative forcing, represent the 90th percentile of projected no-policy emissions as understood nearly two decades ago.

Technological and policy advances since then have revised their likelihood downward: For example, the probabilistic Resources for the Future Socioeconomic Projections (RFF-SPs) indicate estimate a median radiative forcing in 2100 of 5.1 W/m2 (90% credible range of 3.3-7.1 W/m2), with a 0.5% chance of exceeding 8.5 W/m2 and a 1% chance of being below 2.6 W/m2.2

However, this observation is consistent with NCA5. Throughout NCA5 and its supporting materials, RCP8.5 and SSP5-8.5 are identified as the "Very High" scenario, while RCP4.5 and SSP2-4.5 are identified as the "Intermediate" scenario. The report employs a continuum of scenarios, including the “Very Low” SSP1-1.9, the “Low” RCP2.6/SSP1-2.6, and the “High” RCP6.0/SSP3-7.0. Based on the RFF-SPs, the most plausible end-of-century radiative forcing is above that of the Intermediate scenario and below that of the High scenario; the Very Low scenario is even lower likelihood than the Very High scenario.3

Indeed, the amendment itself appears to misstate NCA5 terminology. The amendment repeatedly refers to RCP8.5 and SSP5-8.5 as “high” scenarios when NCA5 reserves the term “High” for RCP6.0/SSP3-7.0 and refers to RCP8.5/SSP5-8.5 as “Very High.” This distinction is important because NCA5’s existing terminology already communicates that RCP8.5/SSP5-8.5 are intended to represent the extreme upper end of the climate scenario range rather than a central value.

2. The amendment incorrectly equates scenario plausibility with policy relevance.

While the amendment correctly states that what NCA5 labels the ‘Very High’ scenario is not “expected, baseline, business-as-usual, likely, [or] central,” it incorrectly also states that it is not “policy relevant.” That conclusion does not follow from climate science, economics, risk analysis, or the scenario literature.

High-forcing simulations are not forecasts of future emissions and their associated climate changes. They are, rather, essential for understanding relationships between temperature change and climate outcomes. For example, NCA5 Chapter 3 relies extensively on warming-level analysis, in which the scientific question is not the climatic consequences of a particular emissions scenario, but how the Earth system responds at particular levels of warming. High-forcing experiments are valuable because they allow models to reach higher warming levels within available simulation periods, providing larger samples for assessing physical responses.

In particular, NCA5 presents projections associated with warming levels of 1.5°C, 2.0°C, 3.0°C and 4.0°C above pre-industrial levels. This range here is similar to the 5th-95th percentile projections for 2100 under the RFF-SPs, which factor into account potential changes in global policy trends,4 and the upper three warming levels are consistent with the range of projections under current policy and technological trends (e.g., Rhodium Climate Outlook, 2025: median of 2.8°C in 2100, 5th-95th percentile range of 2.0-3.9°C).5 Because different full-complexity global climate models have different climate sensitivities, including a very high emissions scenario increases the number of models able to provide information regarding the consequences of plausible high-end warming levels like 4°C.6

The same principle applies in the impacts and economics literature. Damage functions used in integrated assessment models and empirical economic analyses are often calibrated across a wide range of climatic conditions.7 High-end simulations help define the shape of the relationship between warming and damages, especially in the upper tail where nonlinear impacts become important. Eliminating these simulations from consideration would reduce information relevant to understanding high-end but potentially consequential outcomes, including those that could occur under Intermediate or High climate scenarios.

More generally, policy relevance is not synonymous with probability. Flood-control systems, coastal infrastructure, emergency management plans, financial stress tests, and other risk-management tools routinely evaluate outcomes that are not expected but are possible and consequential. The amendment never explains why the National Climate Assessment should abandon this widely accepted principle.

3. High-end climate outcomes remain plausible even though the specific socioeconomic assumptions of SSP5-8.5 are not.

High-end climate scenarios are also more relevant than analyses like those discussed above would suggest because global climate models do not represent and Earth system models do not fully represent positive climate-carbon feedbacks, such as Arctic permafrost thaw, Amazon rainforest dieback, and boreal forest dieback. If these feedbacks are strong, the greenhouse gas concentrations and thus radiative forcing associated with a given emissions scenario will be higher than expected, or than calculated in studies such as those involving the RFF-SPs.8

The amendment also ignores the importance of longer time horizons. The Global Change Research Act directs National Climate Assessments to examine trends over periods extending up to a century. NCA5's Appendix 3 explicitly notes the usefulness of the Very High scenario in considering conditions beyond 2100 because comprehensive post-2100 simulations remain relatively limited. Note that in the new CMIP7 ScenarioMIP scenarios, which the amendment cites as a reference, projected radiative forcing in the H scenario exceeds 8.5 W/m2 within the first half of the 22nd century – these are conditions that can be assessed in scenario runs ending in 2100 only by references to the Very High scenario.9

In short, the implausibility of SSP5-8.5 as a central socioeconomic future does not imply that climate outcomes comparable to those explored under SSP5-8.5 are impossible or irrelevant. Climate outcomes comparable to those of SSP5-8.5 might materialize under lower anthropogenic emissions if climate sensitivity is high or climate-carbon cycle feedbacks strong, and also might materialize within the statutory century timescale of the NCA under more probable high scenarios.

4. The amendment confuses climate forcing scenarios with socioeconomic pathways.

The amendment's treatment of scenario architecture contains a fundamental technical error.

Its discussion of RCP8.5 is broadly correct. RCP8.5 originated in an integrated assessment modeling projection characterized by “high population growth, relatively slow economic development, [and] limited technological advancement.” In the scenario framework, the Representative Concentration Pathway was then decoupled from the generating integrated assessment model – it is Representative because there are multiple different ways to achieve similar forcing pathways.10

Indeed, SSP5-8.5 achieves a similar forcing pathway through a quite different approach. It is built Shared Socioeconomic Pathway 5, “Fossil-Fueled Development”, which does not assume fast population growth or slow economic and technological development. Rather, SSP5 is characterized by relatively slow population growth, rapid technological development, very strong economic growth, and high per-capita energy consumption. The pathway reaches high forcing through rapid economic expansion and fossil-fuel-intensive development, not through the assumptions associated with the socioeconomic pathway used in constructing RCP8.5.11

Conflating the integrated assessment modeling assumptions under RCP8.5 and SSP5-8.5 misses the point: the climate scenarios are not the same as the socioeconomic pathways upon which they are built, and there are multiple different socioeconomic pathways that can lead to high-end climatic outcomes.

5. Section 4 contains vague claims and several clear inaccuracies.

Section 4 attempts to identify “specific NCA5 claim categories that need correction.” However, this section fails at specificity entirely; rather than identifying specific messages of NCA5 that allegedly misconstrue the Very High scenario as an “expected, baseline, business-as-usual, likely, [or] central” scenario, it makes only general (and in many cases erroneous) claims. The amendment therefore provides neither the evidence nor the analysis necessary to justify the report-wide corrections it proposes.

Here are some examples:

As previously noted, the proposed amendment (§4.1) misidentifies NCA5’s “Very High” scenario as the “High” scenario. Section 4.2 notes correctly that Chapter 2 uses the Very High scenario to communicate timing of global temperature thresholds, but does not note that it also does so for the Low, Intermediate, and High scenarios.

As flagged in Sections 4.4 and 4.5, the Energy, Air Quality, and Health chapters do reference literature results derived from the Very High scenarios, but they do so while clearly describing these results as associated with the Very High scenarios. Further, the proposed amendment presents no evidence that the Very High scenario results are key to the NCA5’s assessment messages.

As noted in §4.6, the economics chapter does include damage estimates derived from Very High scenarios, where that is what the literature provides. In particular, Table 19.1b of the economics chapter does include literature results from the Very High scenario, where that is what is present in the literature; it also includes in the same table results for the Intermediate and High scenarios, with the source scenarios carefully marked. The summarize studies accurately reflect the literature and do not bias the chapter’s assessment messages.

Section 4.8 mischaracterizes the federal sea-level rise framework by treating it as though it were an extension of RCP8.5. NCA5's coastal analysis relies heavily on the 2022 Federal Interagency Sea Level Rise Technical Report, which develops a set of planning scenarios spanning from 0.3 m to 2.0 m of global-mean sea-level rise in 2100. These planning scenarios are derived from the full suite of emissions-driven sea-level projections. Figure 9.1 provides a clear cross-walk between these planning scenarios and the IPCC’s emissions-driven scenario. It illustrates that the Low (0.3 m global-mean sea-level rise through 2100), Intermediate-Low (0.5 m through 2100) and Intermediate (1.0 m) sea-level scenarios encompass the IPCC likely ranges for 1.5-4.0°C warming levels, while the Intermediate-High (1.5 m) and High (2.0 m) sea-level scenarios reflect the potential contributions from ice-sheet instabilities. Figure 9.1 also shows that projections through 2050 derived from an extrapolation of observations sit between the Intermediate-Low and Intermediate sea level scenarios.

6. The amendment raises serious procedural and scientific-integrity concerns.

The National Climate Assessment is a Highly Influential Scientific Assessment (HISA), and as such must fulfill numerous requirements under the Information Quality Act, related to criteria including transparency, independence, expertise, and public documentation of review. Gold-standard review, comparable to that undertaken by the National Academies for NCA reports, is an essential part of these requirements. As a HISA, the amended report would also need to fulfill these requirements, but no peer review plan is presented for the amended report.

Further, NOAA’s NAO 202-735D-2: Scientific Integrity requires that NOAA employees “Acknowledge in publications the names and roles of those who made significant contributions to the research”. The proposed amendment is hosted by NOAA but does not acknowledge any authors.

Evidence for the use of a large-language model (LLM) in drafting the proposed amendment (including typical LLM language patterns and document metadata indicating authorship via the Python tool ‘python-docx’) also raises concerns. If a LLM was materially involved in drafting the amendment, USGCRP should explain how the resulting text was validated, reviewed, and verified. Failure to do so would be difficult to reconcile with NOAA's scientific-integrity requirements emphasizing transparency, objectivity, and rigorous review.

Regardless of LLM use, it would be entirely inappropriate for an assessment report developed transparently over multiple years, with a fully disclosed list of authors and having undergone multiple stages of rigorous peer review, to be modified by an anonymous and unreviewed amendment.

Recommendation

In conclusion, the proposed amendment should be withdrawn.

While the amendment correctly observes that RCP8.5 and SSP5-8.5 (the NCA5 Very High scenario) should not be presented as an “expected, baseline, business-as-usual, likely, [or] central” scenario, it does not show that NCA5 does this. Nor does not demonstrate that outputs derived from these pathways are scientifically invalid or policy irrelevant, or that their use to describe the NCA5 Very High scenario biases the key messages of the assessment. The amendment conflates socioeconomic pathway plausibility with climate scenario relevance, contains important factual errors, provides only vague discussion of the findings it seeks to correct, and has not undergone a review process.

USGCRP should instead address updated scenario literature through the development of NCA6. The National Climate Assessment process is designed to evaluate emerging science through transparent literature assessment, multidisciplinary expert review, public comment, interagency review, and independent peer review. That process provides the appropriate mechanism for determining how evolving understanding of future emissions pathways should influence federal climate assessment.

Respectfully submitted,

Robert E. Kopp, Ph.D.
Distinguished Professor of Earth and Planetary Sciences
Rutgers University–New Brunswick

(Affiliation included for identification purposes only. The opinions expressed herein are my own, not those of my employer.)

  1. Richard H. Moss et al., The next Generation of Scenarios for Climate Change Research and Assessment, 463 Nature 747 (2010), http://www.nature.com/doifinder/10.1038/nature08823. ↩

  2. Marcus C. Sarofim et al., High Radiative Forcing Climate Scenario Relevance Analyzed with a Ten-Million-Member Ensemble, 15 Nat Commun 8185 (2024), https://www.nature.com/articles/s41467-024-52437-9. ↩

  3. Id. ↩

  4. Kevin Rennert et al., Comprehensive Evidence Implies a Higher Social Cost of CO2, 610 Nature 687 (2022), https://www.nature.com/articles/s41586-022-05224-9. ↩

  5. [Rhodium Group]{.smallcaps}, Rhodium Climate Outlook 2025: Probabilistic Global Emissions and Energy Baseline Projections (2025), https://climateoutlook.rhg.com/. ↩

  6. Note that the IPCC Sixth Assessment Report assessed very likely range for late-century warming for SSP5-8.5 is 3.3-5.7°C, implying there are some models that provide data regarding the state of the world under 4°C warming only when strongly forced. IPCC, Summary for Policymakers, in Climate change 2021: The physical science basis (V. Masson-Delmotte et al. eds., 2021). ↩

  7. For example, Tamma Carleton et al., Valuing the Global Mortality Consequences of Climate Change Accounting for Adaptation Costs and Benefits, 137 Quarterly Journal of Economics 2037 (2022). ↩

  8. Sarofim et al., supra note 2. ↩

  9. Detlef P. Van Vuuren et al., The Scenario Model Intercomparison Project for CMIP7 (ScenarioMIP-CMIP7), 19 Geoscientific Model Development 2627 (2026), https://gmd.copernicus.org/articles/19/2627/2026/. Note that the new scenarios use a letter-based labeling scheme; the H scenario is the highest of the scenarios. ↩

  10. Moss et al., supra note 1. ↩

  11. Keywan Riahi et al., The Shared Socioeconomic Pathways and Their Energy, Land Use, and Greenhouse Gas Emissions Implications: An Overview, 42 Global Environmental Change 153 (2017). ↩

Updated:

You May Also Enjoy