Assessment status for all CE scenarios and reference models — structured validation across 8 checks using the CE v4.0 decision-reliability framework.
66
Total items
66
Reviewed
0
Not reviewed
0
Needs fixes
100%
Complete
3
Decision-grade
21
Stress-test grade
1
Review required
0
Not validated
How to run a review
Machine checks run automatically — each CE scenario is validated against 5 deterministic arithmetic assertions on every page load (mandate math, horizon years, ceiling consistency, tech-vector sum, abatement margin). Results appear in the Notes panel under AUTO CHECKS.
For the LLM narrative review, click JSON next to any scenario, then click Copy Scenario Validator Prompt and paste both into an external LLM.
Record PASS / WARN / FAIL results for each of the 8 checks in data/reviews/reviews.json and advance the maturity ladder status accordingly.
Key structural model gaps are explicitly disclosed, including UCAP/ELCC, desalination coupling, static interconnection treatment, and excluded SAF demand.
WARN
The top-line emissions arithmetic is close but not cleanly closed: 58.0 Mt × 45% = 26.1 Mt, not 26.0 Mt, and the 2032 mandate path lands at 31.8 Mt versus a stated 32.0 Mt ceiling while several narrative figures use 31.9 Mt.
Tech vectors sum to 27 Mt against a 26 Mt requirement, but the margin is only 1 Mt and the scenario itself discloses that excluded SAF demand of 3.2 GW would erase that margin.
Physical/geographic inputs are mostly region-appropriate for the UAE, but the file mixes UAE and broader GCC framing and includes at least one implausible scalar: years_to_survivability_threshold = 2.5 conflicts with the stated 33.2°C wet-bulb baseline, +0.7°C/decade trend, and narrative that places material exceedance in 2029–2035.
The policy story is directionally coherent for a sovereign-led UAE buildout, but it overstates linkage between the UAE power mandate and EU CBAM exposure on petrochemicals/LNG/aviation, which is not a direct legal non-compliance penalty tied to missing a domestic clean-power target.
FAIL
Fleet and generation figures are internally inconsistent: baseline capacity totals imply 37.4 GW, but mandate_2032 labels all 26.0 GW gas as ccgt_ccus despite the scenario elsewhere stating no CCUS, and the named clean generation math (46.6 + 37.4 + 3.8 = 87.8 TWh) does not reconcile with the target note claiming 84.9 TWh from 16 GW solar at 25% CF and 1.5 GW wind.
The non-compliance cost series is not internally consistent: the target text cites €85/t in 2032, but non_compliance.tax_schedule uses $95/t in 2032, rises to $445/t by 2036 without justification, and the assumption register itself flags the CBAM FX arithmetic as wrong; 38 Mt × €85/t does not support the stated $3.8B cleanly under the documented conversion note.
The schedule is not feasible as written: the scenario requires 10 GW of additional solar plus 3 GW/4h BESS and major district-cooling retrofits by 2032 while simultaneously stating a 15-month financing delay risk, 28% construction productivity loss, a 7-month construction season tightening toward 5 months, and only a 1 Mt abatement margin.
CE USEUse this scenario as a stress narrative for UAE heat-risk and sovereign-financed transition planning, not as an investment-grade CE mandate case. The direction of travel is credible, but the file has blocking inconsistencies in fleet accounting, penalty-cost arithmetic, and delivery timing. Before operational use, reconcile all clean-generation and fleet totals, restate CBAM/non-compliance mechanics on a legally correct basis, and rerun the build schedule with explicit delay and UCAP-adjusted reliability constraints.
Bangladesh Bay of Bengal Climate Transitionbangladesh_bay_of_bengal_transition
The scenario explicitly discloses major structural model gaps, including LOLE/EUE, dispatch endogeneity, fiscal constraints, contract-exit economics, and physical climate risk omissions.
WARN
Mandate arithmetic is not cleanly closed: 40% of the 63.6 Mt baseline is 25.44 Mt (reported as 25.4), and the portfolio reconciliation admits a 0.6 Mt overlap between DSM and gas efficiency while also stating a canonical 26.2 Mt total without fully exposing the underlying five-vector sum.
Abatement coverage is directionally sufficient but not transparently additive: tech_vectors sum to 26.2 Mt only after using DSM at 3.6 Mt, whereas analysis.tech_contributions uses DSM at 4.2 Mt and then subtracts a 0.6 Mt overlap.
The non-compliance series is internally consistent on its own terms (10.7 Mt × $42/t ≈ $0.45B in 2034 and cumulative values sum), but it relies on explicitly low-confidence assumptions about CBAM extension to garments and buyer threshold enforcement, and the scenario mixes €42/t in narrative with $42/t in the tax schedule.
The scenario is mostly geographically grounded in Bangladesh and the Bay of Bengal, but it contains several plausibility flags such as referring to IEC 61400-3 for solar installations and using an offshore-wind-specific standard in a solar note, plus inconsistent JETP dates/amounts ($21.5B in Nov 2023 vs $21.5B in Nov 2022 vs $21.7B in analysis).
The policy story broadly fits Bangladesh's BPDB/BERC/IMF context, but several core drivers are scenario assumptions rather than established legal mandates, especially buyer-imposed grid-intensity thresholds and CBAM extension to garments.
FAIL
Fleet accounting does not balance: baseline_2026 component capacities total 27.83 GW against a stated 26.8 GW, and the mandate path removes 4.8 GW HFO plus 3.2 GW coal while adding intermittent renewables without a demonstrated reliability-adjusted capacity plan; the scenario itself concedes LOLE/EUE and UCAP are unmodeled.
The timeline is not feasible as written for the full portfolio: offshore wind has a 5.0-year permitting timeline and 5.5-year total lead time with first power noted as 2032, leaving almost no schedule slack before the 2033 deadline, while the scenario simultaneously assumes 15 GW of JETP solar operational by 2029 despite 2.5-3.0 year approvals, land constraints, transmission bottlenecks at 88% utilization, and a 3.5-year interconnection queue.
CE USEUse this scenario as a stress-test narrative for Bangladesh's fiscal, trade, and climate-exposure interactions, not as an investment-grade transition plan. The key quantitative weaknesses are fleet inconsistency, weak reliability treatment, and an over-tight deployment schedule for offshore wind and large-scale solar. CE practitioners should preserve the disclosed model-gap framing and policy logic, but rework the capacity balance, timeline sequencing, and abatement reconciliation before using it for formal mandate or financing decisions.
Brazil Amazon Legal Reserve Reforestationbrazil_amazon_reforestation
The scenario explicitly discloses that CE does not model land-use sequestration, carbon-market revenue, non-linear forest growth, and Amazon tipping dynamics, which is the right treatment for a land-use case outside CE's core scope.
WARN
The operative abatement stack sums to 48 MtCO2/yr against the analysis target of 45 MtCO2/yr, but this only works for the net-flux metric and is inconsistent with the headline mandate figure of 517 MtCO2/yr and the listed 89% reduction framing.
The non-compliance schedule is arithmetically consistent on 45 MtCO2 at $22/$30/$40/$55/$75 per t yielding about $0.99/$1.35/$1.80/$2.48/$3.38B, but the mechanism is conceptually mixed because EUDR is a market-access regime, not a per-ton tax, and affected exports are variously stated as €12.8B/yr, $14.5B, and even R$280B ($56B).
The schedule is aggressive: 11.7 Mha must be restored or offset by 2030 while seed supply needs 3 years to scale, smallholders account for 31% of the deficit, and technical assistance plus CAR enforcement capacity appear stretched.
Most physical references fit Brazil's Amazon, but the scenario includes misplaced language such as 'hurricane/cyclone events in the Amazon delta region' and claims 1.5 Mha of degraded/lost mangroves since 1985 despite an intact base of 1.3 Mha, which needs source-tightening.
The Forest Code, CAR, CRA, IBAMA, Amazon Fund, and EUDR pressure channels are broadly coherent, but the scenario overstates direct SBCE penalty exposure while also admitting primary agriculture is excluded from direct mandatory obligations, leaving the enforcement narrative partly self-contradictory.
FAIL
The mandate arithmetic does not close: the target states 89% reduction and 517 MtCO2/yr required reduction from a 2026 baseline, yet the analysis actually uses only 45 MtCO2/yr net-flux change, and the file itself notes that 89% of 580 Mt is 517 Mt while 89% of total 870 Mt would be 774 Mt.
CE USEUse this as a qualitative and fiscal-risk overlay for Amazon land-use transition dynamics, not as a CE-validated abatement case. The core story may be decision-useful for governance, export-risk, and financing analysis, but the mandate definition must be rewritten to one metric only before any CE comparison or ranking. Treat 2030 delivery as contingent on enforcement continuity, seed/TA scale-up, and clarification of how much compliance is reforestation versus CRA offset.
Non-compliance cost arithmetic is internally consistent: the intervention stack totals $5.66B (2.8 + 1.4 + 0.84 + 0.62), the 2026-2030 expected annual losses sum to the stated $9.1B, and the 2030/2035 values match the projections array ($2.7B and $7.4B).
The physical setting is regionally plausible: Phoenix heat, 8.4 GW baseline peak, 118°F event framing, water-power dependence, and Maricopa-specific infrastructure references are geographically consistent.
Key structural model gaps are explicitly disclosed, including cross-domain propagation, non-linear cascade dynamics, and behavioral response, with clear statements that telecom, hospital, and social domains are analog-calibrated rather than CE-native.
WARN
Timeline feasibility is only partially established: while the listed lead times fit before 2035 on paper, the scenario itself flags major execution dependencies such as urban siting for 96 water microgrid sites, AMT/CCI contract execution, NFPA 30 storage variances, and rate-case approvals, so deliverability is not high-confidence.
The governance narrative is not fully clean because the scenario relies on several soft or uncertain mechanisms—especially tower-REIT contracts covering ~95% of macro sites, Starlink fallback for the remainder, and broad multi-agency coordination—rather than a single enforceable resilience mandate across all six domains.
CE USEUse this as a resilience stress-test and cascade-sequencing scenario, not as an emissions or power-transition case; mandate, fleet, and abatement checks are not applicable because the target is reliability under extreme heat, not CO2 reduction. The strongest value is in identifying power as the keystone vulnerability and testing whether Phase 1 milestones by 2028 can prevent the hour-72 to hour-96 hospital failure window. Practitioners should treat the telecom, hospital, logistics, and recovery timelines as scenario-based approximations with stated ±40% uncertainty rather than model-validated forecasts.
China Food Security Under Climate Stress: North China Plain Aquifer Depletion and Yangtze Droughtchina_food_security_climate
Mandate arithmetic closes on the scenario’s own numbers: BAU wheat+r rice in 2030 is 310 Mt versus mandate-path 338 Mt, so the programme adds 28 Mt and lifts self-sufficiency from 88.3% to 96.2% against 351 Mt consumption.
The adaptation vectors cover the stated production-at-risk total, with 25 + 12 + 5 + 8 = 50 Mt protected versus 45 Mt at risk, leaving a stated 5 Mt margin.
The non-compliance cost series is internally consistent: annual path sums to $140B over 2026-2030 and matches the stated two-event expectation and $70B peak annual cost.
Regional inputs are geographically coherent for China: North China Plain aquifer depletion, Yangtze drought, and Heilongjiang-Jilin black soil all align with the named scenario region and there are no obvious misplaced jurisdictional references.
The policy narrative fits China’s governance structure, relying on State Council, MARA, MWR/YRCC, provincial implementation, quotas, reserves, and mandatory programme enforcement in ways that are institutionally plausible.
Key model limitations are explicitly disclosed, especially aquifer irreversibility uncertainty, Yangtze drought-frequency uncertainty, and missing global commodity price transmission within CE.
WARN
The deadline is tight because the scenario still depends on converting 12 Mha of NCP irrigation across highly fragmented 0.6 ha farms, metering 1.2M wells, and delivering 40 Bm³ of added Yangtze storage by 2028-2030, with several benefits hinging on aggressive execution and stated mitigation rather than demonstrated completion.
CE USEUse this scenario as a structured adaptation stress test for Chinese food security and global commodity spillovers, not as a high-confidence implementation forecast. The arithmetic and internal logic are mostly solid, but delivery risk is concentrated in the 2026-2030 execution window for smallholder irrigation conversion and Yangtze water infrastructure. CE practitioners should treat the scenario as decision-useful for vulnerability mapping and contingency planning, with sensitivity testing around adoption rates, permitting slippage, and drought-frequency assumptions.
Colorado River Basin Water-Energy Nexus Crisis Mandatecolorado_river_basin_crisis
Mandate arithmetic closes: 45.0 Mt × (1−15%) = 38.25 Mt, and the mandate path reaches 37.9 Mt in 2032 for a stated 0.35 Mt buffer.
Technology coverage exceeds the target on paper: 5.8 Mt from Four Corners retirement + 5.5 Mt from solar+BESS + 0.5 Mt from Palo Verde uprate = 11.8 Mt versus 6.75 Mt required, even after the revised 2.1 Mt hydro-loss gas backstop risk.
Physical and geographic inputs are largely consistent with the Arizona/Nevada–Colorado River context, and the earlier hydro derate contradiction is explicitly reconciled to 3.4 GW nameplate, 2.2 GW effective, and 1.2 GW lost.
Key CE structural gaps are disclosed directly, including missing hydrology-power coupling, agriculture cascade effects, groundwater substitution, and litigation uncertainty.
WARN
Fleet adequacy is only thinly demonstrated: in the high-load case the reserve margin is 20.99 GW versus a 20.93 GW requirement, leaving just 0.06 GW and depending on the full 1.5 GW of contracted DR plus a simplistic 100% BESS peak contribution assumption.
The non-compliance series is internally reconciled on its own terms (7.9 Mt × rates), but the scenario itself says the carbon-price schedule is a modelled economic proxy rather than an operative AZ/NV legal penalty, so the $95/t to $445/t schedule is not a jurisdictional enforcement cost in the ordinary sense.
The schedule is presented as feasible, but the scenario also lists explicit failure conditions that make 2032 infeasible if ACC approval slips past Q2 2027, if Navajo Nation agreement is not executed by Q4 2026, or if reservoir storage falls below 30%, showing the pathway is highly contingent rather than robust.
The narrative mixes several authorities and scopes awkwardly: a CO2 reduction mandate is justified partly through water-shortage administration, RPS compliance, and NERC TPL planning standards, while also admitting no single federal authority directly mandates utility reserve margin and that the carbon penalty is only a proxy.
CE USEUse this scenario as a stress-case planning narrative for the Arizona/Nevada power-water nexus, not as a cleanly enforceable policy mandate with settled penalty mechanics. The emissions math and technology stack are broadly serviceable, but capacity adequacy is thin under higher load growth and the non-compliance cost stack blends legal penalties with modelled proxies. Practitioners should rerun reserve margin and timeline conclusions with a stricter UCAP/ELCC treatment for solar, BESS, hydro, and DR, and should separate legally operative penalties from scenario-valued economic damages.
European Dunkelflaute Grid Stress Testeuropean_dunkelflaute
Timeline feasibility is uncertain due to the complexity and lead time of proposed hydrogen storage and SMR deployment.
Model gaps identified include correlated multi-country weather events and seasonal storage dispatch, which lack comprehensive mitigation.
CE USECE practitioners should use this scenario to explore risks associated with high renewable penetration without adequate dispatchable backup, especially during Dunkelflaute events. Key limitations include assumptions about the realization of long-duration storage solutions and SMR deployment timelines. Consideration of separate regional constraints is also critical due to model gaps in handling multi-country correlated events.
Green Climate Fund Deployment Gapgreen_climate_fund_deployment_gap
Non-compliance cost: the cost of institutional failure is explicitly and precisely quantified — adaptation gap $193B/yr, L&D need $290–580B/yr vs $800M pledged, GCF approved-to-disbursed ratio 31%, 280 GW locked-in coal in developing nations; these figures are sourced, consistent, and directly usable as CE context parameters
Narrative coherence: the loan-vs-grant distortion ($89.6B OECD-counted vs $30–40B grant-equivalent), accreditation barrier (4.8-yr average accreditation time, designed for World Bank-equivalent institutional capacity), and co-financing requirement exclusion are all internally consistent and analytically important; SIDS and LDC specific data is compelling and well-sourced
WARN
Mandate math: the mandate is institutional (GCF disbursement rate 31%→70%, approval-to-disbursement 4.7yr→18 months, private capital $0.37→$2.00/dollar) rather than CO2-based; required_reduction_mt_co2 = 0; no tech vector CO2 abatement arithmetic exists — the scenario quantifies finance deployment failure, not emission reductions directly
Fleet capacity: not applicable — no power sector fleet data; scenario scope is global climate finance institutions, not energy system transitions
Abatement coverage: no abatement tech vectors; the scenario identifies institutional intervention points (accreditation reform, grant-vs-loan rebalancing, L&D fund scaling, co-financing waiver for LDCs) but does not compute CO2 avoided per intervention; locked-in coal (280 GW) is identified but not mapped to an abatement pathway
Timeline feasibility: GCF has operated for 10+ years at <40% disbursement rate; reaching 70% by 2030 requires governing board consensus among 194 member countries and fundamental accreditation reform — highly ambitious; L&D fund operationalisation ($0 disbursed in 2024) suggests the 2030 scale-up target requires unprecedented institutional acceleration
CE model plausibility: CE fiscal, adaptation, and guidance services can model the economic impacts of adaptation deficits (physical damage costs, fiscal stress from uninsured losses) but cannot model the institutional disbursement mechanics that are the core of this scenario; this scenario provides calibration context for CE analyses, not CE-computable inputs
Model gaps: CE's fundamental inability to model institutional finance is not explicitly documented as a model gap in the scenario; for client-facing CE use, this must be stated: "CE quantifies climate impacts of the finance gap, not the institutional mechanisms causing it"
CE USEQuantitative context and narrative framework for CE analyses involving developing-nation climate finance risk. The $193B/yr adaptation gap, 31% GCF disbursement rate, and 280 GW locked coal are direct CE calibration inputs for fiscal stress, adaptation deficit, and physical damage analyses in developing nations. This scenario CANNOT be run as a standard CE quantitative scenario — use as backdrop documentation. When CE outputs reference developing-nation transition feasibility, cite GCF disbursement constraints explicitly as a finance risk overlay. Pair with NGFS Current Policies and IPCC SSP2-4.5 for full developing-nation inaction-cost framing.
Himalayan Peak Water Transitionhimalayan_peak_water_transition
Non-compliance cost: inaction cost chain explicitly quantified — $28B/yr GDP at risk (wheat + cotton), $1.44B emergency wheat import in cascade trigger year, 180 bps sovereign spread widening, $3.8B capital flight from Karachi markets; inaction cumulative loss $140B by 2050 vs $18B adaptation cost = 7.8:1 cost-benefit ratio; cascade steps clearly linked
CE model plausibility: CE WaterStressService, physical_climate (glacier mass balance), damage, fiscal, and adaptation services all directly applicable; tech vectors explicitly mapped to CE service stack; scenario is well-structured for CE analysis
Narrative coherence: the peak-water masking insight — current floods temporarily INCREASE river flows, creating a false signal of water security that makes adaptation politically impossible — is analytically distinctive and internally consistent; dual extreme (flood NOW, deficit LATER) is the central tension and clearly articulated
WARN
Mandate math: CO2 abatement framework inapplicable — required_reduction_mt_co2 = 0; mandate targets are water-based (25 BCM canal recovery, GLOF warning coverage 100%, post-peak deficit ≤15% of baseline); the cascade model identifies emissions co-effects (+22 Mt/yr emergency thermal + diesel) but these are consequences, not mandate targets
Fleet capacity: 27% hydropower share (1.9 GW of 7.1 GW fleet) is at peak-water deration risk but no fleet_evolution path modeled; hydro PLF decline under peak-water transition unquantified; the reliability window (2030–2040 when floods derate hydro capacity while thermal backup is insufficient) is not resolved as a dispatch adequacy analysis
Abatement coverage: all three tech vector CO2 estimates are 0.0; the 100 Mt/yr indirect emissions benefit of avoiding emergency thermal generation is noted in cascade model but not included in mandate metrics or tech vector accounting — should be documented as a formal co-benefit
Timeline feasibility: canal lining at $400M/yr pace requires 30 years for full completion (explicitly acknowledged); the $18B adaptation financing consortium (World Bank $5B + ADB $4B + GCF $3B + bilateral $3B + domestic $3B) "does not yet exist at the required scale" per the scenario's own analysis notes — the critical operational gap; Pakistan's IMF Extended Fund Facility constraints limit domestic fiscal contribution
Model gaps: only 2 gaps formally documented (peak water timing ±15yr uncertainty, IWT renegotiation probability); key unformalized gaps: (a) groundwater-glacier interaction (aquifer depletion and meltwater decline interact nonlinearly — aquifer loss removes the buffer that makes the post-peak transition survivable); (b) CPEC infrastructure GLOF exposure ($62B of corridors in direct outburst flood paths) unquantified; (c) hydro fleet deration reliability impact
CE USEPakistan Indus Basin water-energy-food security scenario. Use CE WaterStressService with AR6 HKH glacier mass balance projections for near-term (2030–2040) flood and GLOF risk; use post-peak deficit sensitivity bands for 2040–2050 analysis. The 7.8:1 cost-benefit ratio for adaptation is the primary CE decision-support output — valid and well-sourced. Flag adaptation financing gap ($18B consortium not yet assembled) in any client presentation. CE cannot model peak water timing uncertainty (±15 years) — always present 2040 and 2050 sensitivity bands. CPEC infrastructure exposure requires manual GLOF risk overlay.
India Himalayan Glacier Water Stress Transitionindia_glacier_water_stress
Non-compliance cost: penalty is explicitly quantified — $22–34B/yr annual loss (agricultural + hydropower + water supply) with component breakdown; wheat output falls 8–12 Mt, groundwater emergency rationing in 6 states; the range framing ($22–34B) is appropriate given the scenario’s own low confidence rating
CE model plausibility: CE WaterStressService with AR6 HKH glacier projections, physical_climate, adaptation, and fiscal services all applicable; 28 GW glacier-dependent hydropower PLF risk maps to CE grid_stability; India scenario stack is well-supported in CE
Narrative coherence: the binding contradiction — Punjab aquifer running dry at 0.33 m/yr overdraft removes the groundwater buffer precisely as glacier meltwater peaks and declines — is the central analytical insight; dual-buffer degradation (glacier + aquifer simultaneously) is the scenario's distinctive contribution and internally consistent
WARN
Mandate math: CO2 abatement framework inapplicable — required_reduction_mt_co2 = 0; mandate targets are water-based (45 BCM storage, 22 Mha micro-irrigation, aquifer stabilised at 0.05 m/yr); the 100 Mt/yr indirect emissions co-benefit (hydropower preservation offsets coal substitution) is noted in scenario notes but absent from all mandate metrics and tech vector accounting
Fleet capacity: 28 GW glacier-dependent hydropower is the key fleet risk; no fleet_evolution path modeled; PLF maintenance target (42%) is specified but post-peak meltwater PLF decline (how much below 42% without adaptation?) is not quantified; the grid stability implication of losing 35–60 TWh of low-carbon generation annually needs dispatch adequacy analysis
Abatement coverage: all four tech contribution CO2 estimates are 0.0; water metrics are present (93 BCM total storage vs 45 BCM target, +48 BCM margin) but the CO2 co-benefit framework is absent; 100 Mt/yr hydropower preservation co-benefit should be documented as a formal climate mandate co-benefit, not a footnote
Timeline feasibility: 15 BCM new surface storage commissioned by 2028 is implausible — Himalayan reservoir construction typically takes 8–12 years (Tehri: 23 years); the scenario itself rates confidence as "low" and cites ILR political stall (40 years), Sutlej-Yamuna Link interstate dispute (1966–present), and Punjab MSP constraint on crop diversification; the 2028 Phase I milestone requires infrastructure already in advanced construction that is not identified
Model gaps: Sutlej-Yamuna Link dispute (interstate water compact, politically stalled since 1966) is a critical constraint on surface storage deployment but not modeled as a formal constraint; MSP (minimum support price) commitment to wheat/rice is the key barrier to crop diversification but modeled only in confidence rationale, not as a formal structural constraint; run-of-river vs storage hydro dispatch model distinction absent
CE USEIndia Indo-Gangetic Plain water-energy-food security scenario. Use CE WaterStressService with AR6 HKH glacier projections for the 2026–2038 near-term window (higher confidence); flag all 2035 adaptation milestone projections as "low confidence" (scenario’s own rating). The 28 GW glacier-dependent hydropower PLF risk is a direct CE grid_stability input for India scenarios. 100 Mt/yr hydropower co-benefit should be declared in any CE climate output for this scenario. Do NOT cite the 15 BCM Phase I milestone for 2028 without disclosing Himalayan reservoir construction lead time constraints.
Non-compliance cost: the cascade model is the scenario’s analytical core — four explicit steps with triggers, timelines, and dollar estimates; insurer exit ($112B avg cat losses 2021–2025) → mortgage unavailability (4.5% premium-to-value threshold) → property value collapse (35%, $420B) → tax base erosion ($48.4B/yr) → muni bond stress ($242B outstanding); the feedback loop (insurance exit → adaptation funding failure → deeper insurance exit) is well-structured; individual components are credible
CE model plausibility: CE financial_stress, fiscal, land_valuation, and adaptation services map directly to the cascade steps; parametric backstop and FHFA repricing tech vectors explicitly mapped to CE fiscal and financial_stress; the insurance withdrawal cascade is one of the most CE-compatible institutional scenarios in the library
Narrative coherence: the central feedback loop is analytically distinctive and internally consistent; the scenario correctly identifies that Florida Citizens insurer of last resort with 1.3M policies (more than any private insurer) signals complete voluntary market failure — this is accurate and well-sourced; audit flags are embedded transparently
WARN
Mandate math: CO2 abatement framework inapplicable — required_reduction_mt_co2 = 0; the "mandate" is financial system integrity (voluntary market ≥85% coverage, state insurer <$400B exposure); no abatement vector CO2 arithmetic exists; the only emissions connection is that uninsurable assets cannot fund the adaptation infrastructure that reduces physical damage — an indirect feedback, not a mandate metric
Fleet capacity: fleet_evolution = "not_applicable" (explicitly stated); institutional finance scenario with no power generation fleet component
Abatement coverage: tech vectors (parametric backstop, managed retreat, FHFA repricing) are financial system interventions, not abatement pathways; all tech vector CO2 estimates = 0; abatement framework inapplicable
Timeline feasibility: managed retreat scale-up (9,000 → 90,000 buyouts/yr) requires federal direct-purchase authority and NEPA categorical exclusions that do not currently exist; historical 4-yr average property-to-close timeline means the 5-year, 450,000-buyout target requires simultaneous initiation of all properties; voluntary acceptance rate 60–75% means 112,500–180,000 targeted properties stay in risk zone; FHFA 1-year lead time in tech vector conflicts with 18-month rulemaking noted in constraints
Model gaps: (a) California FAIR Plan insolvency risk is the highest-probability systemic event — $460B exposure vs estimated $160B claims-paying capacity, statewide assessment on all CA policyholders if triggered — not formally modeled as a tail risk; (b) cross-market correlation (simultaneous FL + CA + Gulf crisis vs sequential) unmodeled; (c) GSE (Fannie/Freddie) balance sheet impact from repricing 2.4M mortgages not quantified; (d) the $3.4T vs $2.6T total at-risk value discrepancy (AUDIT FLAG HIGH) remains unresolved — use $2.6T as documented floor
CE USEUS physical risk and managed retreat scenario for financial stress analysis. CE financial_stress and fiscal services model the insurance withdrawal cascade directly — this is one of the most CE-computable institutional scenarios. Use $2.6T (documented floor) not $3.4T for total at-risk property value until the $0.77T gap is sourced. Flag the FAIR Plan CA insolvency tail risk ($460B vs $160B capacity) as the highest-probability systemic event not captured in central estimates. The 4-step cascade model is the primary CE client output — uninsurable threshold years (FL 2029, Gulf 2030, CA 2031) are the key scenario dates.
CM: CE WaterStressService (drought_risk parameter with MRC 2019–2020 drought calibration), agricultural damage, fiscal (NPL cascade), and commodity_markets (rice price transmission) all applicable. Dam operating rules documented as HIGH gap with explicit workaround.
NR: Transboundary governance failure framing is analytically distinctive. Tonle Sap lake empirical data (16,000 → 9,800 km², 2000–2023) is observational, not modeled. Laos debt-trap-as-constraint on regional adaptation solidarity is well-articulated. Scenario correctly positions economic threshold as the instrument of diplomatic urgency.
MG: 4 formal model gaps documented — dam operating rules (HIGH, with MRC drought calibration workaround), AWD adoption pace (HIGH, with 600K ha low-case sensitivity), groundwater substitution (MEDIUM, 15% resilience buffer), trade reallocation (MEDIUM, FAO elasticity proxy). Each gap has a documented workaround.
WARN
MM: required_reduction_mt_co2 = 0; mandate is flow-based (≥70% pre-dam dry-season baseline by 2035), not an emissions ceiling. CE mandate validator CO₂ framework is inapplicable. The governing mandate is a diplomatic flow-sharing obligation. Framework mismatch — not an error in the scenario.
FC: 24.6 GW fleet (58% hydro) — but dispatch is dominated by Lancang upstream release decisions that CE cannot model. Dam operating rules not published, subject to commercial and geopolitical discretion (HIGH model gap). Fleet enumeration without dispatch constraints systematically understates dispatch uncertainty across the basin.
AC: All three tech vectors have estimated_mt_co2 = 0 (precision irrigation AWD, mangrove restoration, desalination). Adaptation-only scenario — no decarbonization pathway. CE abatement coverage framework is inapplicable. Note co-benefit: mangrove peat carbon avoided release (8–12 MtCO₂e/yr, ESTIMATED) if restoration target met.
TF: AWD scale-up 8× (250K → 2M ha in 5 years, ~350K ha/yr) flagged as HIGH model gap — no comparable Southeast Asian rice extension programme has achieved this rate. Laos sovereign debt trap (debt service requires maximum hydropower generation) locks out cooperation until 2030. China–LMC binding treaty trajectory rated 'unfavorable' by scenario's own 2025 confidence rationale. Mangrove establishment lag (5–7 years) leaves only 3–5 years of full protective function before 2035 deadline.
CE USEUse WaterStressService drought_risk parameter with MRC 2019–2020 drought as calibration baseline. Apply commodity_markets for rice price transmission (FAO elasticity proxy: 10% supply reduction → 18% price increase). Use CE fiscal service for Vietnam/Cambodia NPL cascade ($4.2B agricultural NPLs → 12% NPL ratio). Critical: do not model Lancang dam release schedules — treat as scenario parameters (not CE outputs). Always run AWD low-case sensitivity (600K ha by 2030 = 60% reduction in modeled water savings). Operative planning assumption: Mekong Delta saltwater intrusion 2.8 Mha by 2035.
Morocco Saharan Green Hydrogen Export Mandatemorocco_green_hydrogen_mandate
MM: Required reduction 8.0 Mt; tech_contributions portfolio total 8.5 Mt (0.5 Mt positive buffer). Portfolio dedup factor 0.649 explicitly calculated with rationale (35.1% cross-vector overlap removed). AUDIT FLAG (MEDIUM) on tech_vector vs tech_contributions discrepancy is transparent — correctly documented as steady-state capacity vs BAU-adjusted annual abatement. Operative mandate math is sound.
AC: Noor CSP+solar (4.2 Mt), Tarfaya/Dakhla wind (2.8 Mt), green H₂ load shifting (1.5 Mt) = 8.5 Mt vs 8.0 Mt required. OCP CBAM exposure addressed via grid decarbonization (412 → 150 g/kWh by 2030). Tech vectors aligned to primary emissions sources. Portfolio dedup documented.
NC: EU CBAM exposure quantified: €3.9B/yr OCP fertilizer penalty + €4.2B/yr RFNBO hydrogen disqualification = €8.1B/yr = 22% of Morocco 2026 GDP. CBAM arithmetic AUDIT FLAG (MEDIUM) transparent — derivation depends on finalized EU CBAM product carbon benchmark (ref: EU Implementing Regulation 2023/956). 2032 EU certification deadline creates hard optionality cliff.
NR: ONEE grid (12.6 GW) vs mandate requirement (23 GW incremental) structural contradiction is clearly stated. OCP–CBAM–hydrogen corridor linkage is analytically distinctive. MASEN deployment track record (Noor I–III on schedule) provides credibility anchor. Dakhla offshore wind + green H₂ co-location as integrated solution architecture is internally coherent.
WARN
FC: Coal (2.1 GW, 10.8 Mt) and gas CCGT (2.4 GW, 4.9 Mt) baseline arithmetic verified in notes. However: no fleet evolution path post-2030; no reserve margin analysis for 52% RE grid at 8.6 GW peak demand; green hydrogen electrolyzer load (5 GW demand) creates new peak load profile not modeled in system reserve margin calculations. 4.5% demand CAGR compounds balancing challenge as firm coal retires.
TF: 6-year horizon (2026–2032) for 23 GW incremental buildout is highly compressed. Critical path: 400 kV transmission backbone Laayoune–Casablanca (1,200 km, MAD 28B) requires 3.5-yr permitting + construction — earliest ready 2029–2030, leaving ≤2 years of network integration before 2032 EU certification. Western Sahara sovereignty dispute creates regulatory uncertainty for projects south of 27.7°N (Dakhla offshore wind, key 1.5 GW). EU green H₂ import prices currently below Moroccan production cost at scale — export economics undemonstrated.
CM: Atlantic wind capacity factor (0.42–0.48) significantly exceeds CE European defaults; mapping_fidelity = 'not_mapped' for Atlantic wind corridor. Saharan solar dust-soiling loss (2.5%/month) not modeled — understates degradation in CE defaults. Noor Midelt Phase 1 PPA ($0.0225/kWh) far below CE cost defaults — CE will overestimate Morocco abatement costs. ONEE single-buyer tariff structure limits IPP financing not captured.
MG: No formal model_gaps section. Key unformalized gaps: (1) Atlantic wind corridor absent from CE geography — no Morocco-specific capacity factor; (2) Saharan dust soiling (~28% annual generation penalty uncorrected); (3) EU green H₂ price trajectory — export viability requires €4–6/kg vs current EU spot below €3/kg (IEA 2025); (4) Xlinks Morocco–UK cable (£21.9B, 10.5 GW, first power 2031) — if delayed, H₂ export scale premise shifts materially; (5) Western Sahara offtake regulatory risk for projects south of 27.7°N not risk-adjusted in projections.
CE USEMorocco has measurable CBAM exposure and a credible institutional actor (MASEN) with a strong delivery track record. Override CE cost defaults with IRENA MENA benchmarks (Atlantic wind LCOE €68–82/MWh onshore; Noor CSP hybrid $0.0225/kWh PPA precedent). Primary CE output: OCP grid carbon intensity trajectory vs CBAM penalty timeline — run annual compliance risk through 2030 as the lead client-facing deliverable. Model two cases: (A) with Xlinks cable 2031 and (B) domestic Morocco-only decarbonization — gap between cases = hydrogen export optionality value. Apply AUDIT FLAG disclosures on CBAM arithmetic in all client reports.
Pakistan Climate-Amplified Flood Risk and Sovereign Debt Stresspakistan_flood_climate_risk
NC: Penalty mechanism explicit and quantified — P90 mega-flood without resilience: $25–35B fiscal shock; IMF primary surplus breach (3–4% GDP); programme suspension; Eurobond default risk. Probability of fiscal crisis from compound climate shock = 35–45% before 2030. Four tech vectors each quantify annual loss reduction: parametric insurance ($0.9B), GLOF EWS ($0.4B), embankments ($0.7B), housing reconstruction ($0.3B) = $2.3B total vs $2.3B target.
CM: CE monte_carlo applicable for annual flood probability trajectory (8%→13%/yr, with explicit annual probabilities in projections). CE parametric_insurance, adaptation, and fiscal services all map to tech vectors. NDMA capacity score (2.8/5.0) and reconstruction efficiency (58%) are usable CE calibration inputs. Critical-path fiscal neutrality models through CE fiscal service.
NR: Binding fiscal neutrality constraint (resilience financed at zero net cost above IMF 2.0% primary surplus target) is analytically distinctive and internally consistent. Pakistan's loss-and-damage framing (not charity — debt-for-nature swaps and parametric products as financial engineering) is well-constructed. 34% probability of at least one mega-flood by 2030 explicitly derived (1−(1−0.10)^4). 2022 event attribution cited (WWA: at least 50% more intense due to climate change).
WARN
MM: required_reduction_mt_co2 = 0; mandate is flood loss reduction from $3.8B → $1.5B/yr; tech contributions sum exactly to $2.3B with zero buffer. Scenario self-rates confidence = 'low'. CE mandate math framework inapplicable for adaptation fiscal scenario — correct framework is climate finance delivery vs IMF surplus constraint.
FC: Scenario records grid_carbon_intensity (480 g/kWh) and annual_emissions (200 Mt CO₂) as context fields only — no power sector mandate, no fleet evolution, no capacity targets. All four tech vectors are adaptation instruments (insurance, EWS, embankments, housing). CE fleet capacity framework inapplicable.
AC: All tech vector CO₂ = 0; adaptation-only scenario. Indirect emissions co-benefit (avoiding emergency diesel generation, crop-loss deforestation, low-quality rebuilding materials) noted but not quantified. CE abatement coverage framework inapplicable — correct accounting is annual flood loss avoidance ($2.3B/yr by 2030).
TF: Scenario self-rates confidence = 'low'. Critical compounding constraints: (1) parametric insurance — NICL/Askari lack capital base for $3B+ annual cover without reinsurance backstop (Swiss Re/Munich Re raising Pakistan premiums post-2022); (2) GLOF EWS scaling 6 → 35 lakes requires Chinese CPEC corridor cooperation in Gilgit-Baltistan — diplomatically uncertain; (3) COP28 L&D fund pledged $775M, disbursed $0 in 2024 — governance structure not finalized; (4) Indus embankment reinforcement (4-year lead time) required a 2026 start to meet 2030 deadline — already time-critical; (5) zero-budget mandate means all four tech vectors depend on a climate finance architecture that does not yet exist at required scale.
MG: No formal model_gaps section. Key unformalized gaps: (1) IMF programme continuation uncertainty (2024–2027 EFF — flagged in confidence rationale only, not formalized); (2) GLOF–monsoon correlation (flood event probability treats GLOF outbursts as independent of La Niña years — both are positively correlated, understating tail risk); (3) L&D fund disbursement timeline vs $2.1B/event NAFIS payout scale ($775M pledged ≪ $2.1B needed); (4) reinsurance backstop availability (post-2022 rate increases by Swiss Re and Munich Re on South Asian cat risk); (5) CPEC resilience infrastructure ($1.4B) not assessed for actual flood protection effectiveness.
CE USEUse CE monte_carlo for annual flood probability trajectory (calibrate to 8%→13%/yr NDMA projections). Apply CE parametric_insurance service for NAFIS payout trigger modeling (flood_extent >15% district crop area). CE fiscal service: model Pakistan's IMF primary surplus constraint vs climate finance offset — the core CE deliverable is whether the financial architecture closes the $2.3B annual loss reduction gap without IMF breach. Always flag confidence = 'low' in client reports. Primary scenario risk: climate finance delivery failure, not climate severity. Note: 34% probability of at least one mega-flood before 2030 is the headline risk framing for institutional clients.
Rust Belt Decarbonization Mandaterust_belt_mandate
AC: Three vectors cover all abatement pathways — solar mega-expansion 50 GW + 10 GW BESS (52 Mt, power generation displacement), oil-fired fleet fuel switch 15 GW conversion + 7 GW retirement (12 Mt, oil-peaker elimination), national building efficiency + AC replacement 8M units (10 Mt, demand reduction). Full sector coverage with no orphaned channel. The oil fleet conversion vector is structurally unique to Saudi Arabia: no other G20 economy can fund its energy transition from the forgone revenue of the assets being retired.
NC: Non-compliance exposure is multi-vector and severe — EU CBAM on $45B SABIC European petrochemical exports ($1.3B/yr CBAM liability for HADEED steel + Maaden aluminum at €85/t); Moody's climate supplement (Oct 2024) flags 2-notch Aa3→A1 downgrade path if NDC non-compliance by 2027 review (+40 bps/notch × $750B outstanding bonds = $3B/yr additional interest per notch); institutional ESG exclusion from sovereign bond funds accelerating post-2024. S&P Climate Watch Negative since June 2024. Total non-compliance CBAM at $200/t (2032 rate): ~$9B/yr across all affected sectors.
CM: CE physical_climate → solar CF applicable (Saudi irradiance 2,285–2,340 kWh/m²/yr; world record $0.0104/kWh Sudair bid context). CE grid_stability → UCAP margin analysis applicable for 73 GW peak + 2.5%/yr growth trajectory. CE scc → abatement value for solar displacement credit. CE fiscal/economics modules handle mandate cost trajectory; saved barrel dividend is separately documented as model gap (CE captures cost side only). Module mapping is the most CE-compatible in the Batch 6 set.
NR: "Burning the furniture" narrative is compelling and internally consistent — Saudi Arabia forgoing $21B/yr in export revenue to power air conditioning is an analytically distinctive policy failure case. Financial paradox (transition is NPV-positive but private project finance markets are structurally unavailable due to 420 bps stranded asset premium) is rigorous and documented. The PIF sovereign co-investment as "necessary unlock" conclusion follows from the financial architecture analysis. Audit flag on solar intensity assumption fully disclosed inline.
WARN
MM: Mandate math shows zero margin — 185 Mt × 40% = 74 Mt required; tech contributions = Solar 52 + Oil fleet 12 + Efficiency 10 = 74 Mt exactly. AUDIT FLAG (HIGH): The 52 Mt solar credit is defensible only under the merit-order assumption that solar preferentially displaces oil-fired peakers (650–700 g/kWh) rather than the fleet-average grid intensity (474 g/kWh). Applying grid average to 92 TWh incremental solar production: 92 TWh × 0.474 = 43.6 Mt — not 52 Mt. Under grid-average intensity: 43.6 + 12 + 10 = 65.6 Mt vs 74 Mt required = 8.3 Mt mandate miss. The merit-order assumption is physically plausible (Saudi daytime solar peak coincides with oil-peaker AC cooling dispatch) but is undocumented and unvalidated in the scenario. With zero margin, any overstatement of solar intensity means the mandate is not met.
FC: Fleet nameplate ample (50 GW solar added to retained ~70 GW gas thermal vs 2030 peak ~80.6 GW). Construction pace is the binding constraint: 50 GW in 4 years = 12.5 GW/yr; stated EPC capacity ceiling is 15 GW/yr (ACWA Power + NEOM + ENGIE in parallel). This leaves only a 2.5 GW/yr buffer — a single contract delay or supply chain disruption removes all slack. Summer construction productivity loss (22%, outdoor labor banned 3.5 hr/day Jun–Sep) means 13 GW/yr must be contracted to deliver 10 GW/yr net. 10 GW BESS is 4-hour duration — covers evening cooling peak hours but not multi-day extreme heat events (Saudi experiences 5–7 day heat waves above 45°C in July–August).
TF: 4-year mandate window (2026–2030) is the tightest in the CE portfolio for a 50 GW deployment. Solar lead time 3.5yr = deliverable only with zero permitting or procurement delays from launch date. AC mandate lead time: 4yr = exactly the mandate deadline; any delivery slippage misses the target. Key binding dependencies: (1) ACWA Power concentrated execution risk — single EPC contractor failure propagates to mandate miss; (2) 14 new 380kV transmission corridors ($8B, 4yr lead) must run in parallel with solar construction; (3) oil-to-gas pipeline for 15 GW of western/southern HFO units not on current gas network (Fadhili GTU Phase 2 expansion required). No schedule buffer on any vector.
MG: Four documented gaps — saved barrel fiscal dividend (high: CE fiscal model captures mandate costs but not the NPV-positive structure where $15B/yr dividend exceeds $10.4B/yr annualized capex, making full mandate NPV positive at >$48/bbl Brent), transmission expansion bottleneck (medium: 14 new 380kV corridors not modeled in CE GridStabilityService), nuclear 2034 gap (medium: final ~15% of 50% renewable target requires nuclear, not deliverable before 2034; scenario correctly scopes to achievable 40% interim milestone), oil thermal UCAP (low: oil steam turbines/distillate peakers have ~0.78 actual UCAP vs 0.85 CCGT proxy used — slight overstatement of 2026 firm capacity).
CE USECE fiscal/economics: mandate cost trajectory (CE captures $52B capex side); separately document $15B/yr saved barrel dividend as NPV-positive offset not in CE fiscal model — this is the primary CE communication gap for Saudi clients. CE grid_stability: UCAP margin analysis for 73 GW 2026 peak + 2.5%/yr growth trajectory (critical path: evening peak coverage with BESS + oil-to-gas conversion). CE physical_climate: solar CF calibration at 2,285–2,340 kWh/m²/yr; apply heat stress 0.76 to construction productivity and panel efficiency degradation at 49°C ambient. Always stress-test solar abatement credit under both merit-order (52 Mt) and grid-average (43.6 Mt) displacement assumptions — 8.3 Mt spread determines mandate pass/fail. Flag zero-margin mandate math as the primary uncertainty for client-facing reports.
Shanxi Province Dual-Carbon Transition Mandateshanxi_dual_carbon_mandate
AC: Four vectors cover all major abatement channels — Loess Plateau wind expansion (5.8 Mt, primary generation displacement), solar PV + pumped hydro (4.2 Mt, peaking displacement + storage enabling), coal CCUS retrofit on 1.9 GW USC fleet (2.5 Mt, firm coal decarbonization), grid efficiency + industrial DERs (1.3 Mt, loss reduction + demand moderation). Tech sum = 13.8 Mt > 13.4 Mt required. 2030 sub-milestone (national carbon peak): mandate_mt_co2 projection = 31.0 Mt = exactly the 31.0 Mt target — sub-milestone met in model. Coal CCUS and wind must be substantially online before 2030 for the sub-milestone to hold.
NC: China ETS compliance mechanism well-documented — current carbon price 98 yuan/t escalating to 200 yuan/t by 2035 on power-sector installations exceeding provincial benchmark (880 g/kWh → 550 g/kWh by 2035). EU CBAM applies to Shanxi specialty steel, coke chemicals, and rare earth oxide exports. Tax schedule verified: Year 1 $30/t × ~8 Mt ≈ $0.24B ✓; Year 2 $46/t ≈ $0.37B ✓. Non-compliance also triggers suspension of provincial new-energy investment licensing and exclusion from national green bond programs — a meaningful operational sanction beyond pure financial cost.
NR: Complex multi-constraint narrative is internally coherent — the entanglement of coal mines with captive power plants (premature plant retirement without mine closure creates liability), Yellow River water allocation competition (coal-mine dewatering vs pumped hydro refill vs CCGT cooling), 2030 national carbon peak sub-milestone as binding intermediate constraint, and cross-border CO₂ pipeline (Shanxi→Inner Mongolia) as second-highest risk item are all logically connected. Audit flag on wind curtailment mandate miss risk is prominent and precisely quantified (5.3 Mt base-case vs 5.8 Mt optimistic = 0.5 Mt swing, determines mandate pass/fail).
WARN
MM: Mandate arithmetic checks out at the headline level — 38.2 Mt × 35% = 13.37 Mt (stated 13.4 ✓); ceiling 24.8 = 38.2 − 13.4 ✓. AUDIT FLAG (HIGH): The tech_contributions table credits wind at 5.8 Mt, which corresponds to the assumption_register "low curtailment" scenario (4% curtailment, SGCC T&D expansion complete by 2028). The scenario's own base-case curtailment is 8% → wind yields 5.3 Mt → total 5.3 + 4.2 + 2.5 + 1.3 = 13.3 Mt < 13.4 Mt required. Under the base case, the mandate fails by 0.1 Mt. The mandate is conditional on SGCC completing the Xinzhou-Taiyuan 500kV corridor by 2028 to reduce curtailment from 8% to 4%. This dependency is documented in the scenario's analysis section but the tech_contributions table presents the optimistic figure without the base-case caveat.
FC: 10.5 GW → 15.0 GW mandate fleet vs 12.79 GW projected 2035 peak demand (9.8 GW × 1.03⁹) = 17.3% nameplate overhead. Resource adequacy unmodeled — continuous-process metallurgy (coke ovens, arc furnaces) and chemical load creates flat 24/7 industrial baseload with very limited demand response. High variable renewable share (wind 4.5 + solar 4.0 = 8.5 GW = 57% of fleet) requires firm backup (USC coal 2.5 GW + gas 0.8 GW + pumped hydro 1.2 GW) and dispatch simulation to validate reliability. Formal LOLE analysis required.
TF: 9-year window is the most comfortable in Batch 6 but faces specific bottlenecks. Wind: 3yr lead → first commissioning 2029 (✓ but tight for 2030 sub-milestone). Solar: 1.5yr → fast. Coal CCUS: 3yr + cross-border CO₂ pipeline 2.5yr Shanxi-Inner Mongolia approval (two provincial governments + NDRC coordination) = longest CCUS approval path in the CE portfolio. Unlike Rust Belt, storage geology is confirmed (Ordos Basin Class I saline aquifer, CNPC 2023 survey) — the risk is political coordination, not geology. Pumped hydro: 4yr lead (NDRC energy plan review + Yellow River water authority sign-off + geological survey) → online ~2030 — required to balance the high-wind portfolio before and during the 2030 national peak sub-milestone.
CM: Three of four tech vectors lack direct CE model entries — wind (capacity proxy only; CE does not capture SGCC curtailment rates, grid absorption constraints, or wind-thermal dispatch correlation), coal CCUS (BECCS proxy: biomass vs. coal post-combustion have different cost curves, energy penalties, and TRL; 20% parasitic load and 85% capture efficiency not in CE BECCS arrays), grid efficiency/DERs (no TECHS_ABATE entry; modeled as T&D loss reduction overlay only). Only solar uses an approximate CE proxy (perovskite utility-scale, noting lower Shanxi CF of 0.19 vs. typical CE calibration). Most model-constrained scenario in Batch 6.
MG: Five documented gaps — resource adequacy for continuous-process industrial load (high), Yellow River water allocation quota (high: pumped hydro + CCGT cooling + coal-mine dewatering compete for fixed provincial quota at 85% utilization; CE has no water stress model for arid inland regions), China ETS carbon price trajectory (medium: escalating 98→200 yuan/t creates material coal retirement incentive effect not captured in CE fiscal model), coal CCUS→BECCS proxy (medium), endogenous coal retirement timing entangled with mine closure (medium: captive power plants, dewatering loads, and stranded mine liability are not separable in CE model). Densest model gap register in Batch 6.
CE USECE ETS compliance cost model: calibrate to China ETS price trajectory (98 yuan/t now → 200 yuan/t 2035) with sector coverage expansion to steel, chemicals, aluminum from 2025–2026. CE grid_stability: resource adequacy for continuous-process industrial load profile (high load factor, limited DR, dominant baseload demand pattern) — required before fleet_evolution.mandate_2035 can be used for reliability planning. CE monte_carlo: model cross-border CO₂ pipeline approval timing (Shanxi-Inner Mongolia; 2.5yr approval path, 30–40% failure probability based on China inter-provincial infrastructure precedents) as the primary CCUS schedule risk. Always present wind contribution under both 8% curtailment (base case: 5.3 Mt, mandate miss) and 4% curtailment (optimistic: 5.8 Mt, mandate met) in client reports — SGCC corridor completion is the binary determining factor. Flag 2030 national carbon peak sub-milestone as the binding intermediate constraint for Dual-Carbon compliance tracking.
South Australia 100% Net Renewable Gridsouth_australia_100pct_renewable
Fleet capacity is consistent with deployment rates and retirement schedules.
Non-compliance cost series is internally consistent and plausible for South Australia.
Physical and geographic inputs match the named region of South Australia.
The policy narrative is consistent with the region's legal and governance framework.
WARN
There is a discrepancy in the emissions reductions estimated for Olympic Dam between tech_vectors and analysis.
The scenario's timeline may be threatened by AEMO certification delays and Olympic Dam electrification timing.
Model gaps noted, such as the lack of differentiation for grid-forming inverter inertia and VPP aggregation economics.
CE USEThis scenario should be used with caution, considering the identified discrepancies and potential timeline risks. Practitioners should account for variations in emission reductions and be wary of model gaps in grid-forming inverter certification and VPP aggregation. The scenario provides a strong narrative for South Australia's transition to a fully renewable grid but relies on critical path timelines and assumptions that warrant close scrutiny.
South Florida Coastal Energy Resilience Mandatesouth_florida_coastal_mandate
AC: Three tech vectors cover all major gas displacement channels — utility solar + 4hr BESS (5.2 Mt: 4.5 GW at 0.26 CF inland Palm Beach/Hendry/Glades counties, displaces gas baseload), Atlantic offshore wind (2.5 Mt: 2.0 GW at 0.42 CF, displaces marginal peakers), grid hardening + C&I microgrids (2.2 Mt: 1.5 GW DERs at <5 MW, bypasses RTO queue). Each vector addresses a distinct dispatch segment. Nuclear 2.2 GW retained as zero-carbon firm capacity throughout — no retirement required. Complete abatement coverage for a 74% gas-heavy mandate grid.
NC: Dual non-compliance mechanism is well-documented — EU CBAM on Florida aerospace and chemicals exports (starts 2036) + proportional clawback of federal IIJA clean-energy block grants. Tax schedule Year 1: stated $0.23B at $32/t implies ~7.2 Mt embedded industrial emissions — plausible for Florida aerospace/chemicals complex. Binary trigger (no grace margin) means partial compliance provides no financial protection. Total 5-year cumulative non-compliance cost exceeds the mandate investment cost for all scenarios — strong compliance incentive. IIJA grant clawback adds a direct public-finance penalty beyond the carbon tax.
NR: Hurricane-coastal framing is internally consistent and analytically distinctive — the mandate couples decarbonization with physical resilience (75% outage reduction per hurricane event from DER undergrounding). CCGT thermal efficiency derate from SST rise (+1.8°C by 2035) is a real and documented mechanism. Three audit flags (zero margin, baseline intensity, methodology switch) are all disclosed prominently. Confidence rated 'low' in scenario — consistent with the complexity of the offshore wind delivery challenge and zero margin.
WARN
MM: Three AUDIT FLAGS compound to create a materially unreliable mandate math section. (1) BASELINE INCONSISTENCY (MEDIUM): grid_carbon_intensity_g_per_kwh = 510 × 42.6 TWh = 21.7 Mt ≠ 19.8 Mt stated; 510 g/kWh is a capacity-weighted gas fleet intensity, not grid-average; operative baseline is 19.8 Mt (consistent with 465 g/kWh grid average). (2) ZERO MARGIN (HIGH): tech sum 5.2 + 2.5 + 2.2 = 9.9 Mt = required 9.9 Mt exactly, margin = 0. Projections show 9.6 Mt at 2035 (0.3 Mt margin) but this contradicts the analysis.estimated_margin_mt_co2 = 0.0. (3) METHODOLOGY SWITCH (MEDIUM): solar abatement uses grid-average intensity (5.2 Mt verified: 4.5 GW × 0.26 × 8,760 × 0.465 = 4.77 Mt ≈ 5.2 Mt ok with CCGT-weighted intensity); offshore wind at grid-average would yield 2.0 GW × 0.42 × 8,760 × 0.465 = 3.42 Mt, not 2.5 Mt, suggesting conservative marginal peaker displacement intensity used for offshore wind without documentation. Methodology switch is defensible but must be explicit.
FC: 9.2 GW → 13.2 GW mandate fleet vs 10.36 GW 2035 peak = 27% nameplate overhead. Two high-severity gaps directly threaten UCAP reliability: (1) resource adequacy/LOLE unmodeled — 13.2 GW nameplate includes 4.5 GW solar at ~0.26 CF and 2.0 GW offshore wind at ~0.42 CF; effective UCAP depends on coincident production during summer peak and hurricane-season availability; (2) hurricane-season capacity derate unmodeled — offshore wind turbines shut down at ~25 m/s (5–15 days per storm track per season); solar output suppressed under tropical cloud cover. Nuclear 2.2 GW (0.92 CF) and retained gas CCGT 3.0 GW provide firm backup but formal LOLE study required before investment-grade planning.
TF: 9-year window but offshore wind creates a structural delivery risk. Atlantic offshore wind lead time = BOEM 4yr + Jones Act vessel scheduling 12–18 months = 5.5yr total → first MW online mid-2031 at earliest. This leaves only 4 years of generation before the 2035 deadline. Jones Act constraints: there are fewer than 6 U.S.-flagged wind turbine installation vessels as of 2026; demand from Northeast, Gulf, and Florida projects creates scheduling congestion. No offshore wind operates anywhere in Florida as of 2026. Given zero margin, a 1-year Jones Act delay pushes the offshore wind contribution to ~2.0 Mt (vs 2.5 Mt for full 4yr) and creates a 0.5 Mt mandate shortfall with no buffer vector. DERs (1.5yr) and solar (2.5yr) have comfortable timelines.
CM: Two of three tech vectors have no CE TECHS_ABATE entry — offshore wind ('not_mapped': no CE v3.7.0 entry; HVDC cable permitting, Jones Act scheduling, hurricane curtailment, marine O&M not captured) and grid hardening/DERs ('not_mapped': distribution-level undergrounding and storm resilience not in CE). Only utility solar uses a CE proxy (perovskite, approximate: FL CF 0.26 vs CE calibration, BESS dispatch optimization not modeled). CE PhysicalClimateService covers heat_stress and flood_risk but not tropical cyclone curtailment (offshore wind hurricane shutdown) or CCGT efficiency derate from SST rise (+1.8°C by 2035 = 3–8% summer peak efficiency loss). Most CE-mapping-limited scenario in Batch 7.
MG: Five documented gaps: resource adequacy/LOLE (high), hurricane-season capacity derate (high), CCGT efficiency derate from SST rise (medium), offshore wind absent from TECHS_ABATE (medium), insurance market exit feedback on project financing (medium — multiple P/C insurers have exited Florida 2022–2025; CE FiscalService does not capture construction financing cost inflation from insurance withdrawal). The combination of two high-severity unmodeled derates (hurricane capacity and LOLE) means the 27% fleet nameplate overhead may be insufficient for peak reliability under compound storm/heat stress — exactly the physical context this coastal scenario is designed to address.
CE USECE solar utility-scale (perovskite proxy): abatement trajectory for 4.5 GW FL deployment at 0.26 CF; BESS dispatch for 4hr firming during evening peak. CE grid_stability: UCAP margin analysis required — flag offshore wind curtailment risk (hurricane-season shutdown) as the primary firm-capacity gap; nuclear 2.2 GW + gas CCGT 3.0 GW is the only reliable UCAP floor. CE physical_climate: apply flood_risk and heat_stress overlays for coastal substation elevation compliance and summer CCGT derate from +1.8°C SST; add tropical cyclone curtailment as manual derate (15 curtailment-days/yr × offshore wind output) pending CE Stage 4 fix. Always present offshore wind contribution as at-risk/contingent given zero margin and Jones Act scheduling risk. Recommend 0.45 GW additional solar+BESS contingency buffer ($0.7B) to eliminate binary compliance exposure — present as cost of insurance, not over-engineering.
Texas AI + Water Stress Compound Scenariotexas_ai_water_stress
NC: Non-compliance penalty is explicitly quantified and severe — ERCOT Uri 2021 precedent: $28B+ economic damage from a single compound event; NERC LOLP >5% triggers federal compliance review; ERCOT energy-only scarcity cap $9,000/MWh reached during shortage; AI hyperscaler SLA penalty $200M+ per operator per 0.001% downtime breach. The 22% LOLP during a P90 12-hour compound event translates to a near-certain multi-hundred-million-dollar economic impact. Penalty framing is scenario-appropriate: water-energy compound events don't generate carbon penalties — they generate economic damage and reliability enforcement actions.
NR: Compound risk framing is analytically rigorous and internally consistent — four stressors are independently documented (P90 drought: 10-year return period, Brazos -32% flow; thermal derate: 9.5% CCGT at 105°F Carnot-derived; wind drought: r = -0.72 correlation with heat domes from ERCOT 2024 study; AI demand: 15 GW continuous + 0.27 BGD water). The compound_event_matrix quantifies the combined effect: 17.2 GW effective capacity loss → 1.4% reserve margin (vs 15% target) → BESS exhaustion at hour 6 → -10.8% post-BESS reserve. Confidence rated 'medium' — appropriate given the wide 1.8–5.1% annual probability range for the joint event. 3.2% annual probability means this event is expected every ~31 years on average (expected more frequently given climate trend).
WARN
MM: CE mandate validator framework inapplicable — this is a compound-risk/reliability scenario, not a CO₂ mandate. The 2% reduction target (99 → 97 Mt) is inherited from the parent texas_ercot_ai_demand scenario and is not the primary objective here. AUDIT FLAG (HIGH, inherited): baseline states grid_carbon_intensity = 340 g/kWh; 340 g/kWh × 490 TWh = 166 Mt ≠ 99 Mt stated. Correct grid-average intensity = ~202 g/kWh. This discrepancy is flagged in the scenario and inherited from the companion scenario. Analysis.estimated_total_mt_co2 = 9.0 Mt (exceeds the 2.0 Mt reduction target) because the tech vectors are modeled as reliability interventions (preventing capacity derate) rather than CO₂ abatement instruments — appropriate for the scenario type but mismatched to CE mandate math framework.
FC: Fleet evolution explicitly marked 'not_applicable' — defers to companion texas_ercot_ai_demand. The compound_event_matrix provides the operative capacity analysis: mandate fleet 136.0 GW nameplate, effective compound-stress capacity 118.8 GW (17.2 GW derated: 4.2 GW cooling, 5.8 GW ambient, 7.2 GW wind) vs P90 peak demand 117.2 GW = 1.4% reserve (far below 15% mandate). AUDIT FLAG (MEDIUM, wind derate): stressor_3 wind depression notes CF drops from 31% → 15.5% = loss of 37 GW × 0.155 = 5.74 GW, not 7.2 GW. The 7.2 GW figure implies CF drops to 11.5%, not 15.5%. At 5.74 GW wind loss, effective capacity = 120.3 GW, reserve = 2.6% — still critically below 15% but LOLP and BESS exhaustion timeline change. Both figures should be disclosed.
AC: CE mandate abatement framework inapplicable — tech vectors target reliability, not CO₂ reduction: closed-loop cooling retrofits (0 Mt CO₂, prevents 3.6 GW capacity derate from Brazos/Colorado basin drought); AI water efficiency (0 Mt CO₂, reduces AI cooling water from 0.27 → 0.03 BGD); long-duration storage 4 GW / 24hr (6 Mt attributed CO₂ — by enabling renewable dispatch to displace gas during non-drought periods); SPP emergency interconnection 3.6 GW (3 Mt attributed CO₂ — import of wind-rich SPP supply). The 9.0 Mt total exceeds the 2.0 Mt mandate target but is driven by the companion scenario's renewable buildout, not this scenario's reliability interventions. Correct assessment framework is reliability (LOLP reduction), not CE mandate abatement coverage.
TF: 5-year window. Cooling retrofits (22 GW, 2.5yr lead): TCEQ permitting + dry-cooling EPC procurement. Feasible within window (✓) but limited Texas-based dry-cooling specialist capacity is a genuine bottleneck. Long-duration storage (4 GW, 3.5yr lead): iron-air commercial scale-up first GW-scale deployment is 2026 — this is a first-of-kind technology deployment within the mandate window; significant commercialization risk. AI water efficiency (1yr lead, ✓). SPP interconnect (4yr lead): the binding constraint is the ERCOT sovereignty statute requiring Congressional authorization — a political barrier CE cannot model and that is historically resistant to change in Texas. At 4yr lead, the SPP interconnect must receive Congressional authorization by 2027 to deliver by 2031, which is the period of maximum ERCOT-federal tension over grid sovereignty.
CM: CE covers portions of this scenario with meaningful limitations — CE WaterStressService drought_multiplier applicable as external capacity derate (manually parameterized); CE GridStabilityService applicable for compound stress reserve margin analysis; CE monte_carlo can model individual stressor probabilities but cannot compute joint CDFs for correlated compound events (the scenario's 3.2% joint probability requires manual combination). AUDIT FLAG: the wind drought/heat dome anti-correlation (r = -0.72) and drought/thermal derate positive correlation are not natively captured in CE — they are hardcoded compound scenario assumptions. CE has no AI data center demand-response model (30% curtailability split between training/inference workloads).
MG: Four documented gaps in the scenario: water-energy nexus feedbacks (high: CE GridStabilityService does not model cooling water availability as a capacity constraint — requires manual derate parameterization), compound event joint probability (high: CE cannot compute joint CDFs for correlated stressors — P90 drought × P90 heat × wind drought × AI demand requires manual combination, stated as 3.2% but uncertainty range 1.8–5.1%), AI demand controllability (medium: 30% curtailability assumption not derived from workload type distribution in CE model), ERCOT scarcity pricing endogeneity (medium: $9,000/MWh scarcity cap investment signals not captured in CE capacity market model — CE is designed for capacity-market contexts, not energy-only markets like ERCOT). Together these gaps mean CE cannot independently validate the 22% LOLP headline figure — it is a scenario-output, not a CE-computed result.
CE USECE WaterStressService: apply drought_multiplier to structural_constraints as external capacity derate (4.2 GW cooling constraint + 5.8 GW ambient derate = 10.0 GW pre-wind). CE GridStabilityService: compound stress reserve margin analysis — use the compound_event_matrix parameters as inputs; validate the 1.4% reserve calculation and confirm BESS duration assumptions (6-hr exhaustion at 14.6 GW continuous draw). CE monte_carlo: model each stressor independently; document joint probability as manually combined (3.2% annual, range 1.8–5.1%), explicitly noting CE cannot natively compute compound CDFs. Primary CE deliverable: show that the BESS + solar mandate fleet is necessary-but-not-sufficient for compound reliability; the 4 GW long-duration storage gap and 22 GW cooling water independence are the policy gaps CE identifies above and beyond the mandate. Always present the Comanche Peak NRC license condition (derate below 40% Squaw Creek reservoir level = 2.4 → 2.1 GW) as a nuclear reliability risk with high visibility for institutional clients. Note: this scenario should be read as a companion to texas_ercot_ai_demand, not as a standalone mandate.
Fleet capacity projections align well with stated deployment rates and retirement schedules for solar and BESS.
The scenario correctly reflects ERCOT's geographic and operational conditions, like grid isolation.
The policy narrative is consistent with Texas's regulatory environment and the role of ERCOT and PUCT.
WARN
Mandate math shows discrepancies: the baseline emissions for the grid intensity are overstated compared to stated CO2 figures, causing a potential over-calculation in abatements.
The technology vectors do not fully address the AI demand surge, creating a potential shortfall in abatement coverage.
Timeline feasibility faces risks due to potential delays in key infrastructure like CREZ 2.0 and landowner opposition to transmission projects.
There are significant model gaps, particularly the absence of endogenous modeling for AI demand growth and capacity planning.
CE USECE practitioners should use this scenario with caution, particularly considering the potential gaps in abatement math due to grid intensity assumptions. While the fleet capacity and region-specific constraints are accurately covered, attention is needed on regulatory and infrastructure risks that can impact project timelines and deployment rates. Dependencies on AI load projections and infrastructure deployment are key limitations.
Vietnam JETP Power Transition Mandatevietnam_jetp_transition
The stated 2030 coverage stack closes on paper because 38 Mt of by-2030 abatement exceeds the 37 Mt requirement and yields a 169 Mt outcome versus the 170 Mt ceiling.
The scenario is geographically grounded in Vietnam-specific assets, institutions, bottlenecks, and transmission topology with no obvious misplaced regional references.
Key structural omissions are explicitly disclosed, including curtailment, coal PPA economics, ammonia co-firing modeling, demand-growth mismatch, and missing UCAP/LOLE adequacy analysis.
WARN
The arithmetic is internally inconsistent across sections: by_2030_abatement is variously described as 38 Mt, as 33 Mt plus 4–5 Mt transmission unlock, and transmission is elsewhere claimed to unlock 12 Mt, so the exact closing math is not singular even though the target is met on paper.
Fleet build-out is only marginally credible because the mandate requires adding 9.0 GW renewables, 3.0 GW BESS, and two 1,800 km 500 kV circuits while cancelling 4 GW under construction and retiring 8 GW in just four years under a system with 6.3-year average permitting and major transmission congestion.
The non-compliance series is not tied cleanly to a defined emissions gap: 55.3 Mt embedded emissions times the stated rates reproduces the annual costs, but the scenario labels these values as JETP/CBAM-like penalties even though the text itself says electronics and textiles are only a forward-looking EU risk rather than current law.
Timeline claims rely heavily on an asserted 3.0-3.5 year JETP fast-track for solar and transmission despite the base permitting averages of 6.3 years for renewables, 8.0 years for offshore wind, and 4.0 years for transmission, so deliverability by 2030 remains uncertain even if offshore wind is excluded from compliance.
The policy narrative overstates legal enforceability in places by calling the JETP milestone 'binding' and linking failure to export penalties and CBAM-style consequences that the scenario itself acknowledges are partly prospective rather than established Vietnamese or EU compliance law.
CE USEUse this as a stress-test narrative for Vietnam's 2030 power-sector peak-emissions constraint, not as an investment-grade build schedule. The core 2030 compliance logic is directionally usable only if practitioners isolate the deliverable stack to solar, DSM, coal actions, and transmission unlock while treating offshore wind, ammonia co-firing, JETP disbursement conditionality, and CBAM-linked export penalties as high-uncertainty overlays. Reliability and contract feasibility remain under-modeled.
Abatement coverage: scenario correctly identifies all three structural VCM failure modes — additionality (REDD+ phantom credits), permanence (reversal risk), and double-counting (absent corresponding adjustments) — and maps each to a specific reform mechanism (CCP label, MRV satellite technology, Article 6.4).
Non-compliance cost: well-characterised — $71T AUM net-zero pledges at risk; SEC greenwashing enforcement, EU CBAM ineligibility for offset-backed claims, and SBTi decertification cascade are all documented regulatory mechanisms.
Narrative coherence: confidence=HIGH is justified by peer-reviewed REDD+ failure documentation (West et al. 2023 Science; Badgley et al. 2022 GCB) and observed VCM volume decline (364→148 MtCO2e, 2021–2024); 'mis-selling scandal' framing is analytically precise.
WARN
Mandate math: not applicable — the 80% CCP penetration target is a market-quality metric, not a national emissions reduction mandate; CE's arithmetic validator finds no baseline emissions to validate; scenario should be treated as a market-integrity stress test, not an emissions abatement scenario.
Timeline feasibility: ICVCM has approved 48 projects as of January 2025 against ~5,000 active Verra VCS projects; achieving 80% CCP-standard penetration by 2030 requires approximately 4,000 additional approvals in 5 years; Article 6.4 first credits expected 2026–2027 with most sovereign negotiations still incomplete.
CE model plausibility and model gaps: CE has no VCM market model, no TECHS_ABATE mapping for any VCM reform mechanism, and no carbon price elasticity model; this scenario functions as institutional and regulatory risk context, not direct abatement pathway modelling; model_gaps entries in JSON are blank — gaps include credit permanence/reversal risk, Article 6.4 political risk, and VCM price elasticity under regulatory regime shift.
CE USEVCM stress scenario — CE has no direct model for voluntary carbon market dynamics. Use CE guidance.py for policy risk framing and ce economics.py for carbon price floor sensitivity. Primary value: institutional risk disclosure — portfolios with offset-backed net-zero pledges face SEC enforcement, SBTi decertification, and CBAM eligibility risk if VCM collapse accelerates. Flag for compliance and institutional strategy clients. Not suitable for CE abatement pathway modelling; use alongside real-economy decarbonisation scenarios to show the correction forcing actual abatement investment.
West Virginia Appalachian Coal-to-Clean Transitionwest_virginia_coal_to_clean
Abatement coverage: tech vectors address all material WV emissions sources — coal power retirement with gas bridge (6.8 Mt, 46%), blue H2/CCUS (4.2 Mt, 29%), ridgeline wind (3.8 Mt, 26%), AML brownfield solar (1.4 Mt, 10%); mine-site solar on 500,000 acres of AML inventory is a distinguishing low-permitting-friction pathway.
WARN
Fleet capacity: Arch2 blue H2/CCUS (DOE $73M demonstration award) is at TRL 6–7, not commercial scale; 4.2 Mt by 2033 requires approximately 1 Mt/yr CO2 injection into Appalachian basin storage — roughly 10× current basin CCS operations; PJM wind interconnection queue averages 5.7yr (2024) which equals or exceeds the remaining mandate horizon for ridgeline wind (3.8 Mt, 26% of target).
Non-compliance cost: political feasibility is the governing constraint — WV has no renewable portfolio standard; governor and legislature actively oppose federal clean energy mandates; 6.8 GW coal stranded assets and ~38,000 coal-dependent jobs create legislative resistance that makes non-compliance structurally more probable than compliance under current political economy.
Timeline feasibility: abatement margin is razor-thin — analysis total 14.7 Mt vs target 14.3 Mt = 0.4 Mt buffer (2.8%); with confidence=LOW and three tech vectors at elevated deployment risk (CCS TRL, PJM queue, gas bridge FERC permitting), any single underperformance eliminates the buffer.
CE model plausibility and model gaps: blue H2/CCUS mapped to BECCS proxy (acknowledged approximation — understates CCS cost and geological siting risk); coal retirement gas bridge and ridgeline wind both map to 'none'; PJM capacity market retirement review not modelled; coal securitization/stranded cost recovery, just transition wage gap, and AML remediation costs all absent from model.
CE USEWest Virginia coal-to-clean scenario. Use CE economics.py for coal stranded asset valuation and carbon risk pricing; use CE damage.py for coal health externalities (PM2.5, black lung). Blue H2/CCUS maps to BECCS proxy — note this understates CCS cost and geological storage risk. PJM capacity market dynamics not modelled — document gap for grid reliability clients. Confidence=LOW with razor-thin margin; recommend using as stress-test outer bound rather than operational baseline. Companion framing: rust_belt_manufacturing_contraction shares PJM and Appalachian political economy context.
Electrification Material Constraint Scenarioelectrification_material_constraint
The abatement coverage is internally consistent: the four tech vectors sum exactly to 480 MtCO2 (95 + 140 + 65 + 180), matching the stated achievable abatement and leaving the acknowledged 200 Mt shortfall against the 680 Mt program requirement.
The scenario explicitly discloses major structural gaps, including missing LOLE/EUE, UCAP/ELCC, hourly balancing, copper price endogeneity, GOES supply, and workforce constraints, which is materially better than leaving these hidden.
WARN
Mandate arithmetic has been mostly repaired, but the file still carries a contradictory high-severity audit note about an obsolete 45% reduction target, creating residual ambiguity even though 4,800 to 3,800 Mt and 680 to 480 Mt now reconcile numerically.
Fleet evolution is directionally coherent but not reliability-grade: the 2033 mandate path shows only 11.0% reserve margin and the bottleneck-constrained path 4.9%, while UCAP, ELCC, and LOLE are explicitly not calculated.
The non-compliance cost series is additive and the 2026-2032 annual values sum to $89.0B, but the file itself says the trajectory is only directionally derived and not independently modeled, so it is not fully traceable.
Physical inputs are mostly US-appropriate, but there are notable plausibility issues such as the malformed decision_windows region string for dw_01 and the tech vector claim of '900,000 Mt/yr' new US copper supply, which is almost certainly a unit error for 900,000 tonnes/yr.
The policy story is broadly US-specific and references IRA, IIJA, DOE, DPA, and NERC appropriately, but some mechanisms are framed too loosely as enforceable mandate structures when they are better understood as programmatic deployment goals and incentive-linked outcomes rather than a single legal mandate.
FAIL
The timeline is not feasible as written because the critical-path mine expansion assumes a 5.0-year total lead time to add supply by 2033 while the scenario also states 19-year average mine development, 12-year US permitting, and severe unresolved litigation and tribal consultation risk at Resolution Copper; that is a blocking mismatch.
CE USEUse this as a directional bottleneck stress test for US electrification rather than as a bankable mandate-delivery plan. The strongest parts are the explicit transformer/copper bottleneck framing and the transparent 480 Mt vs 680 Mt shortfall. Do not use its reliability outcomes or mine-delivery timing for formal planning without external resource-adequacy modeling and a full rework of the mine fast-track schedule and cost traceability.
CE Internal Models
6Drel Labs proprietary models powering the CE platform — shareable with users on request
Theoretical foundation: decision-support scope explicitly defined and disclosed; CE supplements (does not replace) traditional IAMs; additive overlay model (base + shock + policy) is transparent about being a pragmatic approximation, not a general equilibrium derivation
Resolution fitness: country/sub-national and sector-level (6 industry groups) with annual to 5-year timesteps through 2050+; appropriate for CE's decision-support use cases
Operational currency: v4.0 (2026), AR6-informed throughout; scenario engine at v3.7; data sources cited to 2025-2026 institutional publications; actively maintained
WARN
Calibration integrity: shock coefficients (shocks_registry.json, 6 calibrated events) and policy pass-through coefficients (policy_instruments.json, 5 instruments × sectors) are internally calibrated — not externally published or peer-reviewed; periodic recalibration against IEA/IMF published impact estimates required
Uncertainty treatment: Monte Carlo service exists (monte_carlo.py) but not systematically applied across all forecast pipelines; some outputs deterministic; SSP climate ranges used in physical climate services but not propagated as full probability distributions through economic layers
Structural completeness: tipping point dynamics (AMOC, permafrost, ice sheet feedbacks) documented conceptually but not computationally modeled; credit contraction cascade, agent-based behavioral adaptation, and biodiversity-into-production-function feedbacks absent — acknowledged as Phase 3+ roadmap items
Validation record: no external peer-reviewed validation published; backcasting skill not formally tested; internal adversarial review framework (this tracker) provides scenario-level quality control only, not model-level predictive validation
CE USECE is a decision-support platform, not a predictive model. All scenario outputs are directional analysis, not precise forecasts. Platform limitations to disclose to users: (1) shock/policy coefficients are internal estimates — treat as order-of-magnitude; (2) apply manual uncertainty bounds where Monte Carlo is not invoked; (3) tipping point dynamics require manual narrative supplement; (4) reference external IAMs (DICE, NGFS, IEA WEO) for equilibrium price path cross-validation.
CE Balanced Transition Synthesizerce-balanced-transition
Theoretical foundation is grounded in well-established literature, with components sourced from IPCC, NGFS, and CDP.
Calibration integrity is ensured by using authoritative data sources like Munich Re NatCatSERVICE and IMF reports.
Uncertainty treatment is detailed with scenario-specific weight dispersion and calibration uncertainty bands.
Structural completeness is demonstrated by detailed mechanisms and operational equations matching the methodology.
Resolution fitness is appropriate for global, sector-level analysis with company-level calibration.
CE integration alignment is evidenced by NGFS Phase 4 scenario anchoring, ensuring compatibility with other CE models.
Operational currency is maintained with data and scenarios up-to-date as of 2023, reflecting current methodologies.
WARN
Validation record lacks comprehensive out-of-sample testing and could benefit from more historical replay examples.
CE USECE practitioners should use this model for investment decisions focusing on sector pressures, resilience, and opportunities. Its strengths lie in industry-native calibration and its direct applicability to current regulatory frameworks. Users should note the limited validation record and may need to supplement with other data for novel risk combinations not covered in the model.
The model is grounded in well-established theories and extensively references calibration against NGFS, FSB, BoE, IEA, and Swiss Re benchmarks.
The calibration sources are authoritative, current, and well-suited to the model's domain, including NGFS, FSB, and BoE scenarios.
The model's architecture and mechanisms are internally consistent and clearly defined with operational thresholds and equations.
Validation records include numerous historical replays showing strong model accuracy against past events like the European Gas Crisis and Fukushima.
The model's geographic and sectoral resolution is appropriate for its global stress-testing use cases.
The model aligns well with other CE models and accurately states scenario compatibilities, such as with the ECB BES and BoE CBES.
All data sources and methodologies are current, reflecting updates up to 2024 CE environment and scenario families.
WARN
Uncertainty treatment is only partially quantified; while limitations are acknowledged, specific uncertainty bounds for scenarios and outcomes are not fully defined.
CE USECE practitioners should use this model for stress testing scenarios dominated by policy fragmentation and sector fragility, serving as a downside boundary condition. It is particularly strong in identifying non-linear sector fragility and regulatory shock compression. However, avoid using for orderly transition analyses as it overstates pressure under these conditions. Be mindful of the adaptive dynamics and governance influences on fragility outcomes.
The model is grounded in IPCC AR6 and other authoritative sources, ensuring robust theoretical foundations.
Calibration sources such as UNEP Emissions Gap Report 2024 and IPCC AR6 WG3 are authoritative and appropriate.
The model architecture is consistent, with key mechanisms clearly defined and operationally detailed.
The geographic and sectoral resolutions are appropriate for a global, technology-focused model.
The data sources and methodologies are current, aligning with recent reports and technology assessments.
WARN
Uncertainty bounds are acknowledged but not comprehensively quantified, particularly in technology abatement estimates.
The validation record is not explicitly documented; there is no mention of historical data validation or comparable model benchmarks.
Integration with other CE models is noted as misaligned with NGFS policy scenarios, potentially limiting comprehensive analysis.
CE USECE practitioners should use this model to understand the scale of transformative climate solutions needed, especially focusing on technology gaps the model identifies. The model's strengths lie in its unique ability to quantify the 'breakthrough gap,' providing vital insights for technology investors and impact capital. However, caution should be exercised regarding the partially quantified uncertainties in technology deployment and the absence of explicit validation records, which might affect some decision-making scenarios.
CE Transition Opportunity Indexce-transition-opportunity
The model is based on well-established economic theories and principles, with all equations and mechanisms traceable to known literature and calibers.
The calibration sources listed, such as IEA and UNEP reports, are current and authoritative, with explicit acknowledgment of the strengths and limitations of each source.
The model includes a detailed record of validation against known historical events, which contributes to confidence in its forecasting ability.
Its geographic, temporal, and sectoral resolution is robust, providing insights at industry-sector, sub-sector, and company levels.
The model is well-integrated within the CE model family, with clear compatibility with other CE scenarios.
Data sources, scenario families, and methodologies are up-to-date, reflecting current market conditions and projections.
WARN
While key limitations and uncertainties are noted, the model could benefit from deeper quantifications of uncertainty bounds and possible scenario variations.
CE USEThis model should be employed by CE practitioners to identify high-opportunity sectors and companies during the energy transition. Its strengths lie in its detailed opportunity signaling, grounded in robust historical validation, and its integration with broader CE frameworks. However, practitioners should note that while uncertainties and limitations are acknowledged, a more quantified account of these factors could enhance decision-making confidence. The model is most effective when used alongside CE risk models to provide a balanced view of both opportunities and risks within the energy transition context.
CE Physical Hazard Cascade Modelce-physical-cascade
The model's theoretical foundation is well-grounded in compound event literature with published calibration sources.
The calibration uses authoritative sources like Swiss Re Sigma and IPCC AR6, ensuring high calibration integrity.
Structural completeness is demonstrated as the model's architecture aligns with described methodologies.
Validation records show consistent accuracy against historical case studies such as the 2017 California wildfires.
Resolution fitness is appropriate, offering global and region-specific insights for various sectors.
CE integration alignment is clear with precise scenario family support and integration with other CE models.
Operational currency is supported by up-to-date data sources such as CMIP6 and alignment with IPCC AR6 frameworks.
WARN
Uncertainty treatment could be more explicit in terms of bounds and sensitivity analysis, especially for new event types.
CE USECE practitioners should use the CE Physical Hazard Cascade Model to assess compound and cascading climate hazards in regions prone to multi-hazard risks, such as California and the Mediterranean. Its strengths lie in robust validation with comprehensive real-world case studies and integration with sector models for precision analysis. However, practitioners should be cautious of the model's limitations in handling novel event types and the extrapolation of parameters beyond observed ranges.
External Reference Models
25IAMs and benchmark damage models used as CE comparison references
AR6-aligned carbon budgets; cited IPCC AR6 WGIII Ch.1/3; used by ECB, BoE, APRA for stress tests
Highest CE integration alignment of any external reference — NGFS NZ2050 is CE's primary transition benchmark
Phase IV (2023) current; NGFS portal actively maintained with public data downloads
WARN
Damage functions in underlying IAMs (esp. GCAM) understate tail physical risk 3-7× vs Burke meta-analysis
No probability weights by design (central bank policy); within-scenario uncertainty not quantified
Tipping point coverage inconsistent: REMIND-MAgPIE partial, GCAM none; no financial system feedbacks in any IAM
~10-11 global regions only; no country/sub-national outputs for most variables; 5-yr timesteps (no annual)
CE USEPrimary CE transition pathway reference and carbon price benchmark. Use all 6 scenario families (NZ2050, B2°C, DiverNZ, DelTrans, CurrPolicies, NDC) for policy spectrum analysis. For T>2°C or sub-national scenarios, document divergence from NGFS regional aggregates and supplement with physical risk downscaling.
Theoretical foundation: physically grounded; SSP narrative-quantitative framework maps socioeconomic storylines to emissions and climate forcing with explicit IAM ensemble (90+ models, 1,200+ pathways); endorsed by UNFCCC scientific process; AR6 WG3 is the authoritative global mitigation reference.
Uncertainty treatment: ensemble spread explicitly quantified and reported; confidence intervals accompany all key findings; SSP scenario spread (1.9 to 8.5 W/m²) bounds plausible futures; carbon price ranges and technology deployment uncertainties documented with likelihood language (likely/very likely/virtually certain).
Validation record: most extensively peer-reviewed climate science document in existence; 700+ lead authors, 59,000+ review comments, multi-government approval; sequential AR comparisons (AR4 → AR5 → AR6) demonstrate progressive constraint of uncertainty ranges; historical emissions backcasting shows strong skill.
WARN
Calibration integrity: IAM ensemble shows wide inter-model spread — carbon prices for 1.5°C pathways range –,000/tCO₂ by 2030 (>10× spread); individual model parameter sets (damage functions, discount rates, demand elasticities) not systematically validated against empirical microeconomic evidence; BECCS deployment (up to 10 GtCO₂/yr) requires land areas that many IAMs underestimate competing pressures for.
Structural completeness: demand-side transformation systematically underweighted relative to supply-side in many pathways; behavioral and lifestyle change modelled as exogenous assumption rather than endogenous response; equity, distributional impacts, and just transition costs largely absent from formal model structure; tipping point cascade feedbacks (AMOC, permafrost, ice sheet) not represented in standard scenario runs.
Resolution fitness: global and R10-region resolution unsuitable for CE jurisdiction-level analysis; decadal timesteps miss investment cycle dynamics; technology cost curves reflect 2021 data — solar, wind, and battery costs have declined further and faster since cut-off, making near-term (2025–2030) deployment feasibility assessments conservative.
Operational currency: AR6 data cut-off is 2021, published 2022; AR7 in preparation (expected 2027); solar LCOE, EV penetration, and battery storage deployment already departed significantly from AR6 published cost trajectories; CE must supplement AR6 benchmarks with IEA WEO 2025 for any near-term technology deployment assertion.
CE USEAR6 WG3 is CE's primary carbon budget and abatement pathway benchmark. Use SSP1-1.9 and SSP2-4.5 as the CE 'aligned' and 'divergence' reference corridors. Cross-check CE scenario mandate totals against AR6 sector budgets. Supplement AR6 technology cost curves with IEA WEO 2025 for any assertion about 2025–2030 deployment feasibility — AR6 data is 4 years dated for fast-moving sectors. Document explicitly when CE scenario abatement volumes diverge from AR6 compatible pathways (e.g., >50 Mt from a 5% country budget share).
Theoretical foundation: bottom-up energy system model anchored to observed market data; modular sector architecture (power, transport, industry, buildings, fuels) is consistent with engineering-economic analysis; avoids contested theoretical aggregations; widely adopted by G20 governments and institutional investors as primary energy planning reference.
Calibration integrity: calibrated annually to IEA's own proprietary dataset spanning 150 countries; technology cost curves (solar, wind, batteries, hydrogen, EVs) updated to 2024 market clearing prices; capacity addition data directly observed from national statistics — not estimated; IEA unit cost tracking series is among the most granular available for energy technology.
Validation record: annual publication allows systematic ex-post comparison; IEA publicly acknowledges historical underestimation of solar and wind deployment (WEO 2024 acknowledged 2013–2022 systematic under-projection); published corrections demonstrate institutional accountability; WEO track record extensively studied in academic literature; backcasting skill for STEPS scenario broadly reasonable for 5-year horizons.
Resolution fitness: country-level detail for all IEA member and partner countries (70+ countries); technology-level capacity, cost, and investment disaggregation; sub-sector granularity (e.g., steel EAF vs. BF-BOF, passenger vs. freight EV); 2030/2035/2040/2050 milestones with interpolation available; CE primary reference for all fleet_capacity checks.
CE integration alignment: CE fleet_capacity checks directly benchmarked against WEO deployment feasibility ranges; NZE 2050 scenario provides CE's 'decision-support grade' reference corridor; WEO technology cost milestones used in CE abatement cost overlay; grid stability services reference WEO capacity additions for regional power sector analysis.
Operational currency: WEO 2025 represents current-year IEA assessment; energy transition data through 2024; EV sales, solar GW installations, battery costs reflect Q4 2024 actuals; updated macroeconomic assumptions (IMF WEO April 2025 alignment); most current available reference for near-term deployment rate assertions.
WARN
Uncertainty treatment: three-scenario design (NZE/APS/STEPS) represents alternative policy assumptions, not probabilistic uncertainty; no confidence intervals on technology cost projections; demand-side behavior treated as deterministic given policy inputs; WEO 2024 significantly revised near-term gas demand downward from WEO 2023 — systematic demand modelling errors possible across consecutive editions; no explicit probability weights assigned to scenarios.
Structural completeness: energy system focus with GDP growth as exogenous input — macro-economy not endogenous; climate damage feedbacks absent; physical risk impacts (droughts, heat on thermal efficiency, flood damage to infrastructure) not incorporated into demand or supply projections; land use change for bioenergy simplified; NBS and biodiversity feedbacks absent; near-term geopolitical supply disruption risk modelled only in sensitivity scenarios.
Resolution fitness: country-level only for IEA member countries; non-member emerging markets (India, sub-Saharan Africa, Southeast Asia) reported at regional aggregate; subnational resolution not available — CE must use other data sources for jurisdiction-level analysis below national scale.
CE USEIEA WEO 2025 is CE's primary near-term (2025–2035) deployment feasibility reference. Use NZE 2050 capacity milestones as the upper bound for fleet_capacity PASS thresholds in CE scenarios. Flag scenarios where deployment rates exceed WEO NZE pace as fleet_capacity WARN. Use WEO technology cost curves as default CE abatement cost calibration inputs. For jurisdiction-level scenarios below national scale, supplement WEO national data with IRENA, BloombergNEF, or national grid authority data. Always cite WEO edition year — technology cost trajectories shift materially between annual editions.
AR6-updated climate sensitivity; current 2023R version
WARN
Damage function (π₂≈0.00236) implies ~2% GDP loss at 3°C — understates tail risk 5-10× vs Burke meta-analysis
Deterministic SCC; no damage parameter uncertainty propagation
Global aggregate resolution only; unsuitable for CE regional scenarios
FAIL
No tipping point dynamics (AMOC, permafrost, ice-sheet instability)
No financial system feedbacks; no regional disaggregation
CE USESCC lower-bound benchmark (~$220/tCO₂ in 2050) and global damage floor reference only. Do not apply to T>2.5°C or regional scenarios without explicit divergence note.
Theoretical foundation: partial-equilibrium energy system model with CGE elements; recursive dynamic (annual time steps, myopic expectations) avoids the calibration fragility of inter-temporal optimization; energy-water-land-climate coupling is theoretically coherent; multi-sector architecture (energy, agriculture, land use, water, urban) provides the most comprehensive cross-sectoral coupling of any routinely deployed IAM; extensively peer-reviewed since GCAM v3 (2011).
Structural completeness: 32 geopolitical regions, 235 water basins, 18 land types, 12 agricultural commodities; BECCS, DAC, and industrial CCS modelled with technology-level detail; bioenergy-water-food nexus explicitly coupled; novel in GCAM 7: improved industrial sector (material efficiency, circular economy pathways), enhanced hydrogen economy module; among the most structurally complete IAMs available for multi-nexus CE analysis.
WARN
Calibration integrity: core calibration year 2015 (GCAM 7 extends to 2020 for some sectors); key behavioral parameters (demand elasticities, technology adoption logits) estimated from historical literature with wide ranges; energy sector calibrated to IEA data but lags IEA WEO by ~2 years; agricultural parameters calibrated to FAO FAOSTAT through 2018; some near-term (2022–2025) energy transition dynamics pre-dated the calibration — AI data-center load growth absent from baseline.
Uncertainty treatment: GCAM produces deterministic outputs per scenario specification; uncertainty analysis requires manual ensemble or parameter sweep — not built in; carbon price uncertainty spans 3 orders of magnitude across GCAM runs in the AR6 database (C1 scenarios –,000/tCO₂ by 2100); temperature outcome uncertainty not surfaced in standard scenario output; users must independently quantify sensitivity to key parameters (ECS, income elasticities).
Validation record: no systematic backcasting validation published for GCAM 7; GCAM 5–6 backcasting against 2010–2022 energy transition shows overprediction of coal use and underprediction of solar/wind deployment; GCAM 7 incorporates improved solar/wind cost learning curves but comprehensive validation against 2015–2025 realised outcomes not yet published; model improvements described in JGCRI technical notes rather than peer-reviewed journal.
Resolution fitness: 32-region aggregation is too coarse for CE jurisdiction-level use — most regions are multi-country aggregates (e.g., 'Other Asia' aggregates 20+ countries); GCAM lacks subnational resolution; annual-to-decadal timesteps miss investment cycle and policy implementation dynamics relevant to CE's 5-year scenario horizons; CE must disaggregate GCAM regional outputs using country-share proxies for any national-level assertion.
CE integration alignment: GCAM technology pathway outputs (CCS industrial deployment, hydrogen production mix, BECCS potential) referenced by CE scenario feasibility checks; however GCAM's 32-region structure requires manual interpolation to CE jurisdiction level — systematic interpolation errors possible; GCAM's BECCS and DAC trajectories are frequently used as CE MG (model gaps) benchmarks but are known to be optimistic vs. deployment reality; no direct API or data pipeline between GCAM outputs and CE.
Operational currency: GCAM 7.0 released late 2023, open source on GitHub (JGCRI/gcam-core); baseline calibrated to 2020; GCAM 7 represents significant improvement over v6 (hydrogen module, improved industry sector) but AI-driven demand load, 2022–2025 solar overbuild, and rapid battery storage growth are not reflected in baseline trajectories; core scenario database (GCAM-AR6) pre-dates actual 2020–2025 energy transition outcomes.
CE USEUse GCAM 7 outputs for multi-nexus scenario benchmarking only: CCS industrial cluster deployment trajectories, bioenergy potential ceilings, and water-energy-food stress interactions. Do not use GCAM regional outputs directly for CE jurisdiction-level scenarios without country-share disaggregation. For near-term (2025–2035) technology deployment rates, prefer IEA WEO 2025 over GCAM given WEO's more current calibration. GCAM is most valuable for CE scenarios involving long-horizon (2040–2050) land use, agriculture-energy nexus, or multi-sector CDR portfolio analysis. Always document GCAM as the source model when citing technology ceiling or BECCS potential benchmarks, as these are known to be upper-bound estimates.
REMIND-MAgPIEremind-magpie
Potsdam Institute for Climate Impact Research (PIK)
Theoretical foundation: REMIND uses inter-temporal Ramsey-type optimal growth coupled with a detailed energy technology module; MAgPIE provides recursive-dynamic land-agriculture coupling with 200+ grid-cell resolution; combined biophysical and macro-economic feedback is theoretically coherent; NGFS selected REMIND-MAgPIE as one of three backbone models for global net-zero pathway design; extensively peer-reviewed across 50+ publications since 2012.
Structural completeness: REMIND-MAgPIE provides the most sophisticated energy-land-macro coupling of the NGFS backbone IAMs; trade and financial investment allocation modelled endogenously in REMIND; MAgPIE propagates water-land-BECCS constraints into REMIND energy system; industrial sector transformation (electrification, hydrogen, CCS) modelled with technology granularity; carbon budget consistent across sectors via shared CO₂ price signal.
WARN
Calibration integrity: REMIND perfect-foresight optimal-control framework requires a calibrated social discount rate — results are highly sensitive to this assumption; REMIND carbon prices are systematically front-loaded and higher than empirically observed market prices; MAgPIE agricultural parameters calibrated to FAOSTAT 2010–2017; REMIND energy sector calibrated to IEA data through ~2020 with some technology cost updates; AI demand load, 2022–2025 solar overbuild, and rapid battery learning not fully reflected.
Uncertainty treatment: REMIND produces deterministic outputs per scenario specification; NGFS five-scenario spread (Net Zero 2050, Below 2°C, NDC, Delayed Transition, Fragmented World) represents policy assumptions, not full parametric uncertainty; carbon prices vary by 3–4× across equivalent temperature scenarios from different research groups using REMIND; no native Monte Carlo capability; REMIND corner solutions (rapid phase-out of high-carbon assets) common near constraint boundaries.
Validation record: no comprehensive ex-post validation of REMIND 3.x against 2015–2025 energy transition outcomes; REMIND systematically projects high near-term carbon prices (>/tCO₂ by 2030 in Net Zero) that have not materialized; MAgPIE land-use projections vs. observed deforestation rates show mixed skill; model improvements described in PIK technical notes and journals but systematic backcasting not published.
Resolution fitness: REMIND covers ~11 world regions — insufficient for CE jurisdiction-level analysis; MAgPIE 0.5° grid is valuable for land but aggregated to 67 regions for coupled runs; 5-year timesteps miss investment cycle dynamics; perfect-foresight assumption creates implausibly smooth transition paths compared to observed policy volatility; CE must interpolate REMIND regional outputs using country-share proxies.
CE integration alignment: REMIND-MAgPIE is the NGFS Net Zero 2050 and Below 2°C backbone — CE scenarios referencing NGFS alignment should cross-check carbon budget consistency against REMIND outputs; REMIND's social-planner perfect-foresight assumption creates optimistic policy transmission that CE's real-world jurisdiction scenarios should discount; MAgPIE BECCS and bioenergy ceilings directly applicable to CE forestry and NbS scenario feasibility checks; no direct data pipeline between REMIND outputs and CE.
Operational currency: REMIND 3.2 / MAgPIE 4.6 released 2023–2024; NGFS Phase V used earlier REMIND 2.x parameterization; GitHub repositories active (remindmodel/remind, magpiemodel/magpie); SSP calibration baseline 2019 with selective updates; solar/wind cost learning rates in recent REMIND runs improved but 2024–2025 overbuild trajectory not captured; IIASA ScenarioMIP contributions for AR7 include REMIND updates in progress.
CE USEUse REMIND-MAgPIE for CE carbon budget consistency checks and NGFS alignment assertions. REMIND Net Zero 2050 and Below 2°C carbon price trajectories set the upper bound for CE carbon price assumptions in high-ambition scenarios. MAgPIE bioenergy potential ceilings (sustainable biomass supply) directly applicable as BECCS and forestry abatement upper bounds in CE NbS scenarios — treat as optimistic and apply 20–30% discount for real-world land tenure constraints. Do not use REMIND regional outputs directly for jurisdiction-level CE analysis without country-share disaggregation. Flag CE scenarios where abatement trajectories deviate >2× from REMIND Net Zero 2050 technology deployment pace as timeline_feasibility WARN.
Theoretical foundation: MESSAGEix uses linear programming energy system optimization with a macro-economic module (MACRO/DMAGIC); GLOBIOM couples agricultural and land-use dynamics with global bioenergy and BECCS constraints; one of the oldest continuously developed IAMs (IIASA, 1970s origin); SSP2 marker scenario developed using MESSAGE; NGFS backbone model; theoretical basis documented across decades of peer-reviewed IIASA publications; Python/GAMS implementation (message_ix) well-documented.
Structural completeness: MESSAGE covers hundreds of energy technologies with detailed resource constraints; GLOBIOM integrates 18 land-cover types, livestock, crop systems, and bioenergy; food-energy-water nexus formally coupled; comprehensive GHG accounting (CO₂, CH₄, N₂O, F-gases); industrial sector (steel, cement, chemicals) represented with technology-level detail; trade flows modelled; among the most technologically granular IAMs for bioenergy and BECCS pathway design.
Validation record: SSP2 marker scenarios (MESSAGE is the SSP2 benchmark) extensively validated and peer-reviewed; IIASA publishes detailed methodology reports and scenario documentation; historical energy system comparisons vs. IEA data published; 50+ year institutional history at IIASA provides accumulated validation record; MESSAGEix open-source release (GitHub: iiasa/message_ix) enables independent reproducibility checks.
WARN
Calibration integrity: MESSAGE energy system calibrated to IEA historical data through ~2015 for core model with selective updates to 2020; GLOBIOM agricultural parameters calibrated to FAOSTAT and MODIS satellite data through ~2018; bioenergy yield and land carbon stock parameters drawn from literature with wide uncertainty bands; near-term solar/wind cost trajectory updates inconsistent across scenario variants; discount rates and social cost assumptions vary by user group, reducing cross-run comparability.
Uncertainty treatment: linear programming structure produces deterministic outputs — small parameter perturbations near constraint boundaries can yield discontinuous results (corner solutions); NGFS scenario spread represents policy assumptions rather than parametric uncertainty; bioenergy deployment extremely sensitive to land availability constraints that are uncertain; no native Monte Carlo framework; IIASA scenario ensembles require manual construction.
Resolution fitness: MESSAGE covers 11 world regions in core configuration; GLOBIOM covers 37+ regions for coupled runs; insufficient resolution for CE jurisdiction-level analysis; EU treated as single region; 5-year timesteps; IIASA provides some country disaggregation for selected variables on request but not in standard model output; CE must interpolate regional MESSAGE outputs.
CE integration alignment: MESSAGE-GLOBIOM is the NGFS SSP1 (sustainability) backbone — CE scenarios targeting sustainable development alignment should cross-check against MESSAGE outputs; MESSAGE's detailed bioenergy and BECCS technology representation directly applicable to CE abatement ceiling checks for forestry and industrial CCS; no direct CE data pipeline; MESSAGE's SSP-based economic trajectory assumptions may diverge from CE's jurisdiction-specific context, particularly for emerging markets.
Operational currency: MESSAGEix-GLOBIOM 2.0 is current; core NGFS Phase V parameterization uses earlier calibration vintage; GitHub repository (iiasa/message_ix) actively maintained; GLOBIOM updated to ~2022 for some scenario variants; IIASA ScenarioMIP contributions for AR7 in progress with MESSAGE updates; technology cost trajectories lag IEA WEO 2025 by 2–3 years for fast-moving sectors (solar, batteries, EVs).
CE USEUse MESSAGE-GLOBIOM for NGFS SSP1-alignment cross-checks and bioenergy/BECCS potential benchmarking. MESSAGE's 400+ energy technology representation is CE's most detailed reference for industrial decarbonization technology pathways (hydrogen, CCS, electrification). GLOBIOM bioenergy supply curves directly applicable as upper bounds for CE NbS and forestry abatement ceilings. For SSP2-aligned scenarios, MESSAGE is the authoritative benchmark — CE should document deviations from MESSAGE SSP2 energy trajectory. Do not use MESSAGE regional outputs for jurisdiction-level analysis without disaggregation. Cross-reference MESSAGE and REMIND carbon price ranges when CE scenario carbon price assumptions exceed /tCO₂.
IMAGE (Integrated Model to Assess the Global Environment)image-pbl
Theoretical foundation: IMAGE uses a systems-dynamics recursive-dynamic approach — avoids the calibration fragility of inter-temporal optimization while maintaining physical-biogeochemical coherence; among IAMs, IMAGE uniquely integrates energy-economy, land use (CLUE-S), vegetation dynamics (LPJmL), water cycle, biodiversity (GLOBIO), and detailed ecosystem services in a single coupled framework; used in IPCC AR6 scenario database; 40-year development history at PBL Netherlands.
Structural completeness: IMAGE's defining strength is multi-system completeness — land-use change model (CLUE-S allocation), natural vegetation dynamics, carbon cycle (terrestrial + ocean), biodiversity module (GLOBIO), food system, water cycle, and ecosystem service quantification are all explicitly modelled and coupled; NbS abatement potential, reforestation carbon sequestration, and land-use change emission factors are the most physically grounded available in any routinely deployed IAM; 0.5° gridded land output provides unprecedented spatial detail for ecosystem analysis.
CE integration alignment: IMAGE is CE's primary benchmark for NbS (nature-based solutions) abatement potentials and land-use change emission factors; IMAGE's CLUE-S land allocation provides the most physically rigorous forest regrowth, restoration, and avoided deforestation estimates available from an IAM; CE forestry and NbS scenario abatement ceilings should reference IMAGE upper bounds; IMAGE GLOBIO biodiversity projections directly applicable to CE natural capital scenarios; PBL Netherlands provides accessible summary datasets aligned with Dutch government policy reporting.
WARN
Calibration integrity: IMAGE 3.3 (2020) energy system calibrated to IEA data through 2017; land and agriculture parameters calibrated to FAO 2016; ecosystem parameter calibration based on MODIS land cover and global forest inventory datasets through 2018; technology adoption curves are scenario-specific assumptions rather than fully calibrated behavioral parameters; near-term energy system trajectories do not reflect 2020–2025 solar/wind cost overbuild; systematic recalibration cycle slower than GCAM or REMIND.
Uncertainty treatment: IMAGE recursive-dynamic structure is deterministic per scenario; ecosystem-scale uncertainties (carbon sink strength, land carbon stocks, NbS permanence) are large but not probabilistically propagated; GLOBIO biodiversity uncertainty not formally bounded; IMAGE scenario spread in AR6 database shows moderate inter-scenario variation; tipping point dynamics (AMOC weakening effects on European agriculture) present conceptually in IMAGE but not computationally coupled to economic layers; NbS abatement permanence risk not captured.
Validation record: IMAGE SSP1 scenarios (van Vuuren et al. 2017) validated against observations through 2015; land-use change projections compared against MODIS satellite data with mixed skill; NbS abatement estimates partially cross-checked against field measurement literature; energy system backcasting less rigorous than GCAM or MESSAGE; PBL institutional accountability is high (Dutch government advisory), but systematic backcasting against 2015–2025 outcomes not comprehensively published for IMAGE 3.3.
Resolution fitness: IMAGE energy-economy uses 26 world regions — insufficient for CE jurisdiction-level analysis; IMAGE land module operates at 0.5° globally (high spatial resolution for land), making it uniquely valuable for country-level NbS potential estimation; 5-year timesteps for energy-economy, annual for land; CE can directly use IMAGE gridded land outputs for NbS potential disaggregation to national level, but energy system outputs require regional interpolation.
Operational currency: IMAGE 3.3 released 2020; AR6 IPCC database uses IMAGE 3.2 (2018) for many scenario contributions; IMAGE 3.3+ updates in progress for AR7 ScenarioMIP; land-use calibration relatively stable (land change is slow), but energy cost trajectories lag IEA WEO 2025 by ~5 years; IMAGE development cycle slower than GCAM/REMIND; most recent comprehensive IMAGE methodology paper 2023 (IMAGE 3.3 update); PBL government mandate ensures continued maintenance.
CE USEIMAGE is CE's primary NbS benchmark model. Use IMAGE 3.3 gridded land outputs for country-level NbS abatement potential ceilings — disaggregate IMAGE 0.5° outputs to national boundaries for CE scenario abatement_coverage upper bounds on reforestation, avoided deforestation, peatland restoration, and mangrove rehabilitation. Apply a 25–40% discount to IMAGE NbS potential estimates for real-world land tenure, governance, and permanence risk. IMAGE energy system outputs (26 regions) require interpolation before CE jurisdiction use. For biodiversity co-benefits of CE nature scenarios, IMAGE GLOBIO projections provide the most rigorous available estimate. Do not use IMAGE for near-term (2025–2030) technology deployment rates — use IEA WEO 2025 instead.
FUND 3.9 (Framework for Uncertainty, Negotiation, and Distribution)fund-3
Theoretical foundation: FUND uses calibrated damage functions mapping global mean temperature change to sector-specific welfare impacts across 16 world regions; SCC is computed as the marginal welfare cost of an incremental ton of CO2 via integrated climate-damage-welfare chain; multi-sector coverage includes agriculture, forestry, water, energy, sea level rise, ecosystems, health (malaria, diarrhea, cardiovascular), and extreme weather; one of the three US Interagency Working Group (IWG) SCC models alongside DICE and PAGE; extensively published by Tol since mid-1990s.
Uncertainty treatment: FUND has been used for more probabilistic SCC analyses than any comparable damage model; Monte Carlo assessments over damage function parameters, equilibrium climate sensitivity, and economic baseline published across multiple US IWG reports (2010, 2013, 2016, 2021); full SCC distribution routinely reported; distributional uncertainty over income elasticity of damage adaptation formally characterized; FUND probabilistic track record is substantially stronger than most IAMs.
WARN
Calibration integrity: FUND damage functions calibrated to econometric estimates and epidemiological studies predominantly from 1990s-2000s literature; agricultural adaptation modelled as smooth cost-minimizing adjustment that overstates real-world adaptive capacity; key parameters (temperature sensitivity thresholds, income elasticity of adaptation) carry large uncertainty and have not been systematically recalibrated against post-2010 attribution science; FUND systematically produces lower SCC estimates than DICE or PAGE, partly due to optimistic agricultural adaptation and net-positive warming effects at low temperatures in some sectors.
Structural completeness: FUND is a damage/welfare model, not an integrated assessment model — no energy system, technology deployment, or mitigation pathway representation; tipping points absent from FUND 3.9 standard formulation; financial contagion, stranded assets, and supply chain disruption not modelled; ecosystem damage simplified to monetized biodiversity loss; FUND cannot be used for mitigation pathway design or technology deployment benchmarking; structural coverage limited to damage cost estimation.
Validation record: damage function calibration is inherently difficult to validate out-of-sample over multi-decadal horizons; Tol (2009, 2014) meta-analyses provide cross-study comparison but not out-of-sample validation; FUND agriculture damage predictions partially validated against observed crop yield literature but contested; FUND sea level rise damage estimates debated (Tol vs. Anthoff et al.); FUND low SCC estimates vs. DICE/PAGE remain subject to academic critique regarding damage function choices; no systematic post-2015 backcasting published.
Resolution fitness: 16 world regions — insufficient for CE jurisdiction-level analysis; no country-level disaggregation in standard model; temperature-to-damage mapping assumes smooth functions without sub-regional heterogeneity; annual timesteps; CE cannot directly apply FUND regional outputs to country-specific scenarios; FUND sectoral damage breakdowns (by damage type) are more useful to CE than aggregate welfare outputs.
CE integration alignment: FUND provides CE a lower-bound SCC reference and sector-specific damage type breakdown; FUND SCC estimates (~-50/tCO2 central at 3% discount rate) represent the conservative anchor for CE carbon pricing assumption ranges; FUND adaptation cost modules inform CE adaptation scenario economic framing; FUND health damage estimates (malaria, heat stress, diarrheal disease) applicable to CE physical risk scenario narrative calibration; FUND is not applicable for mitigation pathway, technology deployment, or NGFS alignment analysis.
Operational currency: FUND 3.9 is current stable version; development significantly slowed — last major structural update circa 2013-2014; R package available (dmbptol/fund on GitHub); US IWG 2021 SCC update evaluated FUND alongside DICE and PAGE with some parameter refreshes; significant post-AR5 academic criticism of FUND damage functions; FUND is no longer considered state-of-art for structural damage modeling but remains policy-relevant as regulatory SCC lower-bound benchmark; no major FUND 4.0 successor announced.
CE USEUse FUND 3.9 exclusively as a lower-bound SCC reference and sectoral damage decomposition tool. FUND central SCC estimates (~-50/tCO2 at 3% discount, 2020 USD) represent the regulatory conservative anchor; CE scenarios with carbon pricing assumptions below FUND central SCC should flag non_compliance_cost as WARN. FUND sectoral damage type breakdown (agriculture vs. sea level rise vs. health vs. energy vs. ecosystems) is directly useful for CE physical risk scenario damage attribution. FUND Monte Carlo SCC distributions applicable to CE uncertainty treatment for carbon pricing scenarios. Do not use FUND for mitigation pathway design, technology deployment benchmarking, or NGFS alignment checks — use REMIND, MESSAGE, or GCAM for those functions.
PAGE 2020 (Policy Analysis of the Greenhouse Effect)page-2020
Theoretical foundation: PAGE uses a welfare-economic framework mapping emissions to radiative forcing to global mean temperature to sector-specific monetised damages; unique among the IWG triad in explicitly incorporating discontinuity/tipping-point parameters as stochastic abrupt-change events; designed from inception for policy appraisal under deep uncertainty; SCC computed as marginal present value of incremental CO2 damage; underpinned Stern Review (2006) and Cambridge Judge energy policy work; extensively documented by Hope in peer-reviewed literature; PAGE20 formally updated calibration from PAGE09.
Uncertainty treatment: PAGE is the most probabilistically native IWG model — all major parameters represented as probability distributions rather than point estimates; Monte Carlo sampling is standard across all PAGE outputs rather than an add-on; tipping point (discontinuity) parameters formally included with probability and magnitude distributions; equilibrium climate sensitivity uncertainty propagated end-to-end; PAGE routinely produces full SCC distribution including tail risk; fat-tail behavior makes PAGE particularly valuable for CE risk scenario characterisation beyond expected-value analysis.
WARN
Calibration integrity: PAGE parameter distributions calibrated to literature ranges — appropriate for probabilistic analysis but creates very wide SCC ranges ( to >/tCO2 under fat-tail draws); key damage exponents and sector damage fractions calibrated to older literature with significant expert judgment; economic growth baselines follow generic developing/developed country trajectories not explicitly SSP-calibrated; PAGE 2020 updated key parameters vs. PAGE09 (higher central and upper SCC) but core damage function architecture unchanged since 2009; high PAGE SCC estimates reflect calibration choices about damage severity and discount rate that remain contested (Stern vs. Nordhaus debate).
Structural completeness: PAGE is a damage/welfare assessment model, not an IAM; no energy system, technology deployment, or mitigation pathway representation; 8 world regions — coarser than FUND or any IAM; damage sectors include sea level rise, economic market, non-economic non-market, and discontinuity/abrupt change; PAGE lacks sector-specific breakdowns (health, agriculture, water separately) that FUND provides; natural capital accounting and biodiversity absent; PAGE cannot be used for anything beyond SCC and aggregate damage estimation.
Validation record: PAGE high SCC estimates (Stern Review +/tCO2) and later PAGE09/20 central values extensively debated in academic literature; Nordhaus-Stern discount rate controversy has obscured what is calibration-driven vs. discount-rate-driven SCC variation; tipping point representation difficult to validate empirically; PAGE20 published parameter review against AR6 literature; Cambridge working paper documentation thorough; however PAGE has received significantly less peer scrutiny of damage function specifics than FUND; no systematic backcasting of PAGE temperature-damage outputs against observed impacts published.
Resolution fitness: 8 world regions — the coarsest geographic resolution among reviewed models; no country-level detail; temperature-damage functions do not vary spatially within regions; sea level rise damages aggregate across all low-lying coastlines within each region; CE jurisdiction-level scenario design cannot directly use PAGE regional outputs; annual timesteps; PAGE is appropriate only for global SCC estimation and aggregate fat-tail risk characterisation.
CE integration alignment: PAGE provides CE the upper-bound SCC estimates and the primary fat-tail damage perspective for risk scenarios; PAGE explicit tipping point (discontinuity) parameters directly applicable to CE tipping point scenario narrative framing; Stern Review PAGE SCC estimates form the UK government and many EU regulatory SCC reference; CE high-ambition and tail-risk scenarios should document alignment with PAGE SCC upper range; PAGE tipping point damage distributions applicable to CE catastrophe scenario design; PAGE not applicable for mitigation pathway, technology deployment, or NGFS alignment analysis.
Operational currency: PAGE 2020 is current; Cambridge active development; however PAGE remains a niche SCC model without active IAM community maintenance infrastructure; PAGE-ICE (tipping points integration) extension under development at Cambridge; Python implementation available; US EPA 2023 SCC update evaluated PAGE alongside DICE and FUND with PAGE 2020 parameters; key calibration assumptions still debated in post-AR6 literature; PAGE 2020 updates Stern Review parameters but core 2009 architecture unchanged; no PAGE 3.0 major revision announced.
CE USEUse PAGE 2020 as CE upper-bound SCC reference and fat-tail damage anchor. PAGE central SCC (-200/tCO2 at 3% discount, 2020 USD) provides the high-ambition carbon pricing reference; PAGE 95th-percentile estimates (-2000+/tCO2) bound CE extreme tail-risk scenarios. PAGE tipping point (discontinuity) parameter distributions directly applicable to CE scenario design for abrupt climate change narratives — reference PAGE mean discontinuity damage fractions when constructing CE tipping point scenario parameters. CE scenarios invoking catastrophic or tail-risk damage narratives should cite PAGE 2020 as supporting reference. Do not use PAGE for mitigation pathway, technology deployment, NGFS alignment, or sector-specific damage decomposition — use FUND for the latter and REMIND/MESSAGE/GCAM for the former.
World Bank Country Climate Development Reportsworld-bank-ccdr
Structural completeness: CCDRs uniquely provide country-level integration of physical climate risk, macro-fiscal impact, poverty and inequality effects, sector-specific vulnerability (agriculture, energy, water, coastal zones), adaptation investment cost estimates, and policy recommendations in a single country-specific package; the policy-to-impact chain is more complete than any single IAM for CE jurisdiction-level scenario purposes; higher-quality CCDRs (Vietnam, India, Brazil, Egypt) include energy transition pathway analysis, Just Transition components, and fiscal sustainability assessment; coverage explicitly designed for development-finance decision-making.
Resolution fitness: CCDRs are the highest-resolution external reference available for CE jurisdiction-level scenario design — country-specific analysis aligns directly with CE scenario structure; fiscal cost estimates, adaptation investment needs, and sector vulnerability assessments are directly applicable to CE mandate_math and non_compliance_cost check calibration; select CCDRs include subnational disaggregation (provinces, river basins, coastal zones); no interpolation or disaggregation required from global model outputs; CE should cross-reference the relevant CCDR for every country-focused scenario.
CE integration alignment: World Bank CCDRs are the most directly CE-applicable external reference for jurisdiction-specific scenario calibration; fiscal risk estimates and adaptation cost ranges from CCDRs directly support CE abatement_coverage and fiscal scenario notes; CE workflow should flag when a relevant CCDR exists for the scenario jurisdiction and cite it in verdict_notes; CCDR sector vulnerability assessments (energy transition costs, agricultural adaptation needs, coastal protection costs) directly populate CE physical risk scenario parameters; CCDR investment gap estimates provide CE upper-bound adaptation cost references.
Operational currency: CCDRs published 2022-2025 for ~40 countries; coverage actively expanding — World Bank targeting 60+ countries by 2026; most recent CCDRs represent current World Bank analytical state-of-art; key CCDRs relevant to CE: Vietnam (2022), India (2022), Brazil (2023), Egypt (2024), Indonesia (2023), South Africa (2023), Bangladesh (2022), Pakistan (2023), Morocco (2022); reports freely available at World Bank Open Knowledge Repository; typically refreshed every 3-5 years; current series is the first and most analytical cycle.
WARN
Theoretical foundation: CCDRs are an analytical framework combining multiple external models rather than a unified model — theoretical coherence varies by country report; physical climate scenarios drawn from NGFS and CMIP6; macro-economic impacts from country-specific CGE models (often ENVISAGE or national models) or reduced-form assessments; damage estimation uses varied methods per country; the 2022 CCDR methodology guidance establishes a meta-framework but individual reports vary significantly in analytical depth and model choice; framework designed for policy advisory context, not research-grade modeling.
Calibration integrity: calibration quality depends entirely on which models are used in each country CCDR; World Bank ENVISAGE CGE uses Penn World Table and GTAP data; physical climate inputs from CMIP6 downscaling vary by country and modeling center; country economic baseline from World Bank Development Indicators (vintage varies); freshness of calibration data inconsistent across reports — some 2022 CCDRs use 2018-2019 economic baselines; no single calibration standard enforced across the CCDR series; quality of land-use and agricultural vulnerability calibration especially variable.
Uncertainty treatment: CCDRs typically present scenario ranges (high/baseline/low) rather than formal probability distributions; sensitivity analysis provided in methodology appendices of higher-quality CCDRs; physical climate uncertainty addressed via NGFS scenario spread; economic uncertainty treatment inconsistent across the series — some CCDRs use formal Monte Carlo for fiscal risk, others use qualitative banding; formal uncertainty quantification not standardized; tail-risk and tipping point treatment absent from most CCDR macro-economic assessments.
Validation record: CCDRs are policy advisory documents subject to World Bank peer review and government counterpart review, not research-grade models with formal validation standards; no systematic ex-post validation framework applied to CCDR economic projections; accuracy of fiscal cost and adaptation investment estimates not systematically tracked against outcomes; quality ranges substantially across the series — major economies (Vietnam, India, Brazil) receive more rigorous analytical treatment than smaller countries; CCDR methodology guidance published but adherence varies.
CE USECCDRs are CE primary jurisdiction-level reference. For every CE scenario targeting a country with an available CCDR, that report should be consulted and cited. Use CCDR adaptation investment cost estimates as the reference range for CE adaptation scenario abatement_coverage upper bounds. Use CCDR fiscal risk assessments to calibrate CE non_compliance_cost and fiscal scenario parameters. Use CCDR sector vulnerability findings (energy, agriculture, water, coastal) to populate CE physical risk narrative details. When CE scenario assumptions deviate significantly from CCDR findings (e.g., lower damage estimates, higher adaptation capacity), flag as model_gaps WARN and document the basis for deviation. CCDRs do not replace IAMs for global carbon budget or technology pathway analysis — use NGFS backbone models for those functions. Available CCDRs: Vietnam (2022), India (2022), Bangladesh (2022), Morocco (2022), Brazil (2023), Indonesia (2023), Pakistan (2023), South Africa (2023), Egypt (2024) and 30+ others.
TF: PPP-weighted aggregate demand, inflation expectations anchoring, and Climate Transition Risk haircut module are all theoretically well-grounded. Sovereign risk premium and trade channel adjustment mechanisms are standard international economics.
CI: April/October revision cycle ensures calibration against latest data. IMF Article IV bilateral surveillance provides independent country-level validation. WEO track record for 1-2yr forecasts is well-documented.
SC: All six declared scenario types (Reference, Downside, Climate Fragmentation) are fully specified. Debt sustainability overlay and sovereign spread mechanisms are complete structural components.
RF: Country-level resolution with sector decomposition is appropriate for portfolio macro baseline. 195-country coverage with PPP weighting matches CE's cross-border portfolio scope.
IA: IMF WEO is the primary macro baseline for CE's economic model suite and NGFS macro anchor. Direct compatibility with NGFS Phase IV orderly transition scenario calibration.
OC: April 2026 edition reflects latest policy developments including US tariff regime changes, European CBAM implementation, and emerging market debt stress updates.
WARN
UT: Scenario-based uncertainty representation (Reference / Downside / Upside) without probabilistic confidence intervals. Fat-tail macro risk (e.g., debt cliff, cascade sovereign default) is captured via scenarios only — probability weighting is judgement-based.
VR: Climate Transition Risk module introduced in 2022 WEO — insufficient track record to validate against observed policy fragmentation outcomes. Core macro forecasting performance is well-established; transition-specific mechanisms are not yet backtested.
CE USECE uses IMF WEO as the macro baseline floor anchoring all economic model outputs. Primary use cases: sovereign fiscal constraint assessment, NGFS macro scenario calibration, and emerging market growth trajectory for cross-border portfolio analysis. Disclose to users: (1) transition-specific mechanisms are newer and less validated than core macro forecasting; (2) uncertainty bands are scenario-based not probabilistic — avoid treating the Reference case as a central forecast with known confidence intervals.
TF: Rational expectations formation, financial conditions index transmission, and sectoral investment elasticity are all theoretically well-grounded. Term structure dynamics and green bond transmission channel are recent additions consistent with current financial economics literature.
CI: Fed CCAR/DFAST calibration cycle ensures alignment with latest US macro and financial data. FRB/US parameter estimates are updated with each stress test cycle — among the most frequently recalibrated macro models.
SC: All declared mechanisms (financial conditions, labor dynamics, investment elasticity, expectations anchoring) are fully operationalised. Balance sheet dynamics and green finance channel are complete structural components.
RF: US-level sectoral resolution with investment elasticity differentiation is appropriate for US portfolio financing cost and capex transmission analysis.
IA: FRB/US feeds directly into US regulatory stress test scenarios used in DFAST and CCAR — direct alignment with US financial sector counterparty risk analysis.
OC: v2025 reflects post-2022 rate cycle dynamics and incorporates updated financial conditions index parameters from the recent tightening period.
WARN
UT: Uncertainty expressed through Fed stress test scenario set (Baseline / Adverse / Severely Adverse) — not probabilistic. Tail scenarios are expert-designed, not generated from a probability distribution.
VR: Climate-specific extensions (green bond channel, transition finance transmission) are recent additions without a full calibration cycle of backtesting against observed green/brown financing differentials.
CE USECE uses FRB/US for US monetary policy transmission analysis: how interest rate changes propagate to sector-level clean energy financing costs, corporate capex, and household demand. Also the reference model for US financial sector counterparty analysis under climate stress scenarios. Disclose to users: (1) US-centric — requires NiGEM coupling for non-US jurisdictions; (2) physical climate risk is externally specified, not endogenous — cannot generate compound physical-financial stress scenarios from first principles.
NiGEM Global Macro Modelnigem-global
National Institute of Economic and Social Research (NIESR)
TF: Multi-country bilateral trade linkages, exchange rate dynamics, and commodity price endogeneity are theoretically well-grounded. Sovereign spread dynamics and CBAM scenario are recent additions consistent with current international trade and finance literature.
SC: All declared cross-border mechanisms (bilateral trade flows, commodity price transmission, capital flow reversal, trade fragmentation) are fully operationalised across 50+ country models.
RF: 50+ country resolution with bilateral trade matrix is the appropriate granularity for cross-border shock propagation analysis in CE's multi-region portfolio context.
IA: NGFS Phase IV cross-border scenario alignment is explicit — NiGEM is the designated cross-border propagation model for NGFS scenario analysis.
OC: 2025 calibration incorporates US tariff escalation dynamics, EU CBAM implementation data, and post-COVID supply chain restructuring patterns.
WARN
CI: Bilateral trade matrix calibration uses WTO data with a 12–18 month reporting lag — rapidly evolving supply chain fragmentation (US-China tech decoupling, EU-Asia CBAM dynamics) may not be fully reflected in current parameter estimates.
UT: Scenario-based uncertainty representation without explicit probability weighting of trade fragmentation and CBAM retaliation outcomes. Fat-tail trade war scenarios are constructed by expert judgement.
VR: Green technology spillover mechanism is structurally motivated but has not been backtested against observed technology diffusion episodes (e.g., solar learning curve international propagation). Mechanism validity is theoretical.
CE USECE uses NiGEM for cross-border shock propagation analysis: CBAM impacts on Asian export competitiveness, critical minerals supply disruption propagation, and delayed-transition policy divergence between regional blocs. Disclose to users: (1) domestic US financial sector transmission is less detailed than FRB/US — use NiGEM + FRB/US in combination for US cross-border + domestic analysis; (2) trade matrix calibration lags may understate current supply chain fragmentation effects.
TF: Multi-model ensemble architecture, SSP scenario mapping, carbon cycle feedbacks, and regional pattern amplification are all theoretically sound and represent the international scientific consensus. Directly underpins IPCC AR6 — highest theoretical legitimacy.
CI: CMIP6 calibration is the result of international coordinated climate model intercomparison — PCMDI performance metrics provide systematic cross-model validation.
SC: Full SSP1-1.9 through SSP5-8.5 scenario coverage with carbon cycle feedbacks, tipping point probability, and regional pattern amplification all operationalised.
RF: Global 50–100 km ensemble provides appropriate resolution for long-run sector-level physical risk analysis; facility-level requires downscaling or GFDL supplementation.
IA: Direct alignment with IPCC AR6, NGFS Phase IV, and TCFD physical risk framework — the highest institutional integration of any climate model in CE's library.
UT: 40+ model ensemble with explicit spread quantification provides the most rigorous deep uncertainty representation available — probability distributions across SSP pathways are well-characterised.
WARN
OC: CMIP6 AR6 vintage (2020–2022); CMIP7 models are in preparation. Some structural improvements to ocean carbon sink representation and cloud feedback parameterisation have been identified since AR6. CMIP6 remains the regulatory standard but is not the latest-vintage scientific state.
VR: Tipping point probability estimates (AMOC weakening, ice sheet instability) carry very wide uncertainty and cannot be observationally validated on model evaluation timescales. These outputs should be treated as indicative probability ranges, not calibrated estimates.
CE USECE uses CMIP6 as the scenario-conditioned physical risk backbone for long-run portfolio analysis and NGFS/TCFD regulatory disclosure. Primary use cases: SSP-conditioned hazard projections 2035–2100, physical risk sector overlay, and NGFS scenario alignment for regulatory reporting. Disclose to users: (1) CMIP6 AR6 vintage — newer model developments not yet incorporated; (2) for facility-level precision, supplement with GFDL process credibility layer for ocean-driven and compound hazards; (3) tipping point probabilities are indicative ranges, not calibrated estimates.
TF: ERA5 reanalysis methodology (4D-Var data assimilation with 80+ years of observation integration) is the state-of-the-art for observed climate state reconstruction. Theoretical foundation is the ECMWF operational analysis system — highest international credibility.
CI: Continuous update cycle with latest observational data; ECMWF validation reports provide systematic performance documentation. Temperature and large-scale circulation calibration is excellent.
SC: Historical baseline, trend detection, operational disruption correlation, near-term projection, and physical-to-financial bridge are all fully operationalised structural components.
RF: ERA5's 31 km global grid (0.28°) provides sufficient resolution for sector-level physical risk; downscaling available for facility-level assessment.
IA: ERA5 is the observational anchor for CMIP6 historical period validation and the physical-to-financial bridge for Munich Re / Swiss Re loss calibration — direct integration with CE's physical risk architecture.
OC: ERA5 is updated monthly with the latest reanalysis data — most current observational climate record available.
WARN
UT: Reanalysis uncertainty is low for temperature but higher for precipitation extremes (especially convective events) and tropical cyclone intensity — both key drivers for insurance sector physical risk. Users should apply wider uncertainty bands for these specific hazard types.
VR: Physical-to-financial bridge (ERA5 event → corporate operational loss) is company-specific — quality of loss calibration varies by sector, geography, and data availability from individual companies. No universal validation standard exists for this bridge.
CE USECE uses ERA5-calibrated for near-term operational physical risk, company-specific loss calibration, and extreme event attribution. Primary use cases: 1–5yr physical risk for operational disruption analysis, backtesting company loss models against historical events, and climate attribution support for regulatory disclosure. Disclose to users: (1) ERA5 is not scenario-conditioned — cannot project under different emissions pathways; (2) physical-to-financial bridge calibration quality is company-specific and should be disclosed; (3) for precipitation extremes and tropical cyclone intensity, apply wider uncertainty bands than for temperature-driven hazards.
TF: Coupled ocean-atmosphere dynamics, high-resolution hydrology, sea level rise from ice dynamics, compound event simulation, and land-atmosphere teleconnections are all physically well-grounded and represent the state-of-the-art in process-based earth system modelling.
CI: GFDL CM4/ESM4 is extensively validated in PCMDI CMIP6 performance metrics; consistently ranks in the top tier for ocean heat content, hydrology, and tropical cyclone simulation. Swiss Re Sigma compound event calibration provides financial validation layer.
SC: All declared mechanisms are fully operationalised including compound event co-occurrence, ocean heat content vertical mixing, and urban/coastal morphology effects.
RF: 25 km regional configuration provides facility-level resolution capability — the highest resolution of any CE climate model for coastal and hydrological hazard assessment.
IA: CE uses GFDL as the process-credibility anchor for CMIP6 ensemble tail validation and insurance catastrophe tail calibration — direct integration with CE's physical risk architecture.
VR: 40+ years of hurricane track and intensity validation; FEMA NFIP flood insurance claims calibration; IPCC AR6 WG1 Ch9 citation for ocean and cryosphere components — strongest validation record of any CE climate model for its specific hazard domains.
WARN
OC: CM4/ESM4 vintage 2019/2020; incremental parameter updates published but no major model release since. GFDL next-generation development ongoing — some cloud and convective parameterisations may have been improved in interim versions not yet fully integrated into CE's GFDL pipeline.
UT: Small ensemble size (limited compute budget for regional high-resolution runs) means uncertainty quantification is narrower than CMIP6 ensemble. GFDL process credibility compensates within its core hazard domains, but tail uncertainty for extreme events beyond the historical record is less well-characterised.
CE USECE uses GFDL as the process-credibility anchor for specific hazard types: tropical cyclones, storm surge, compound flooding, hydrological drought, and sea level rise beyond 2050. Primary use cases: insurance tail calibration reference, long-duration infrastructure asset risk for coastal and hydrology-sensitive sectors, and CMIP6 ensemble tail validation. Disclose to users: (1) limited SSP coverage (SSP2-4.5, SSP3-7.0, SSP5-8.5 only); (2) narrow ensemble — uncertainty quantification is best complemented with CMIP6 ensemble spread; (3) raw outputs require expert post-processing — not plug-and-play.
TF: Capacity factor modelling from empirical site surveys, curtailment as a function of transmission capacity, BESS integration shifting, LCOE on silicon learning curve, and JETP fast-track are all theoretically and empirically well-grounded for Vietnam-specific conditions.
SC: All declared mechanisms are operationalised including grid frequency stability premium, water-energy nexus interaction, and module degradation tropics adjustment.
RF: Province-level resolution (Ninh Thuan, Binh Thuan, Central Highlands) is appropriate for project-level investment analysis and JETP pathway quantification.
IA: PDP8 scenario alignment ensures compatibility with Vietnam's official planning framework. JETP IPIP pathway parameters are the most current available.
VR: EVN/NLDC curtailment data (2019–2025) provides robust empirical validation of the curtailment model. GWEC site survey data validates capacity factor ranges.
WARN
CI: Post-Circuit-4+5 curtailment reduction (32%→14%) is modelled from grid studies, not confirmed operational data. Actual post-commissioning curtailment may differ from transmission study assumptions.
OC: BESS procurement framework (EVN tender Q4 2026) is pending — if delayed, BESS cost parameters and curtailment reduction projections may need revision.
UT: Sub-provincial capacity factor variance is not modelled — project-level yield uncertainty requires PVGIS site-specific overlay. Post-FIT merchant price risk is not quantified.
CE USECE uses this model for Vietnam solar LCOE benchmarking, JETP pathway analysis, and curtailment risk assessment. Disclose to users: (1) curtailment post-Circuit-4+5 is modelled, not operational — validate against commissioning data when available; (2) merchant price risk for post-2030 projects is not captured; (3) for site-specific analysis, supplement with PVGIS data overlay.
TF: Capacity factor from met-ocean surveys, monopile CAPEX structure, MND clearance timeline mapping, typhoon resilience premium, and FIT economics are all theoretically well-grounded and Vietnam-specific.
SC: All declared mechanisms operationalised including foundation design sensitivity, typhoon resilience CAPEX, offshore collector system allocation, and OEM supply chain economics.
RF: Site-level resolution (Ba Ria–Binh Thuan corridor, Binh Thuan coast) is appropriate for project-level feasibility and JETP pilot tranche analysis.
IA: PDP8 offshore wind annex targets and JETP Article 7 FIT parameters are fully aligned. MND clearance timeline is the most accurate permitting model available.
WARN
CI: Capacity factors derived from Fugro/Ørsted met-ocean survey models — not confirmed by operational turbine data. No COD project exists in Vietnam to validate against. First-of-kind risk applies.
OC: Vingroup turbine JV parameters are MOU-stage ($72M/MW, 2027); actual unit economics may differ. Vietnamese monopile fabrication (Vung Tau) not yet operational — supply chain cost curve is forward-projection.
UT: Grid stability limits for >5 GW offshore injection are not modelled — frequency and voltage stability constraints could bind before LCOE limits and are not captured in current cost-risk framework.
VR: MND clearance fast-track under JETP Article 7 has no executed precedent — timeline assumption based on SOP documentation and expert assessment, not completed clearances.
CE USECE uses this model for Vietnam offshore wind LCOE analysis, JETP Article 7 pilot tranche risk assessment, and MND clearance timeline modelling for Ba Ria–Binh Thuan sites. Disclose to users: (1) no operational Vietnamese offshore project — first-of-kind risk applies to all parameters; (2) Vingroup OEM supply chain is MOU-stage not contractual — cost curve may be optimistic for 2026–2028; (3) grid stability limits for large-scale offshore injection are unstudied; (4) MND clearance fast-track has no executed precedent.
Theoretical foundation: decision-support scope explicitly defined and disclosed; CE supplements (does not replace) traditional IAMs; additive overlay model (base + shock + policy) is transparent about being a pragmatic approximation, not a general equilibrium derivation
Resolution fitness: country/sub-national and sector-level (6 industry groups) with annual to 5-year timesteps through 2050+; appropriate for CE's decision-support use cases
Operational currency: v4.0 (2026), AR6-informed throughout; scenario engine at v3.7; data sources cited to 2025-2026 institutional publications; actively maintained
WARN
Calibration integrity: shock coefficients (shocks_registry.json, 6 calibrated events) and policy pass-through coefficients (policy_instruments.json, 5 instruments × sectors) are internally calibrated — not externally published or peer-reviewed; periodic recalibration against IEA/IMF published impact estimates required
Uncertainty treatment: Monte Carlo service exists (monte_carlo.py) but not systematically applied across all forecast pipelines; some outputs deterministic; SSP climate ranges used in physical climate services but not propagated as full probability distributions through economic layers
Structural completeness: tipping point dynamics (AMOC, permafrost, ice sheet feedbacks) documented conceptually but not computationally modeled; credit contraction cascade, agent-based behavioral adaptation, and biodiversity-into-production-function feedbacks absent — acknowledged as Phase 3+ roadmap items
Validation record: no external peer-reviewed validation published; backcasting skill not formally tested; internal adversarial review framework (this tracker) provides scenario-level quality control only, not model-level predictive validation
CE USECE is a decision-support platform, not a predictive model. All scenario outputs are directional analysis, not precise forecasts. Platform limitations to disclose to users: (1) shock/policy coefficients are internal estimates — treat as order-of-magnitude; (2) apply manual uncertainty bounds where Monte Carlo is not invoked; (3) tipping point dynamics require manual narrative supplement; (4) reference external IAMs (DICE, NGFS, IEA WEO) for equilibrium price path cross-validation.
CE Balanced Transition Synthesizerce-balanced-transition
Theoretical foundation is grounded in well-established literature, with components sourced from IPCC, NGFS, and CDP.
Calibration integrity is ensured by using authoritative data sources like Munich Re NatCatSERVICE and IMF reports.
Uncertainty treatment is detailed with scenario-specific weight dispersion and calibration uncertainty bands.
Structural completeness is demonstrated by detailed mechanisms and operational equations matching the methodology.
Resolution fitness is appropriate for global, sector-level analysis with company-level calibration.
CE integration alignment is evidenced by NGFS Phase 4 scenario anchoring, ensuring compatibility with other CE models.
Operational currency is maintained with data and scenarios up-to-date as of 2023, reflecting current methodologies.
WARN
Validation record lacks comprehensive out-of-sample testing and could benefit from more historical replay examples.
CE USECE practitioners should use this model for investment decisions focusing on sector pressures, resilience, and opportunities. Its strengths lie in industry-native calibration and its direct applicability to current regulatory frameworks. Users should note the limited validation record and may need to supplement with other data for novel risk combinations not covered in the model.
The model is grounded in well-established theories and extensively references calibration against NGFS, FSB, BoE, IEA, and Swiss Re benchmarks.
The calibration sources are authoritative, current, and well-suited to the model's domain, including NGFS, FSB, and BoE scenarios.
The model's architecture and mechanisms are internally consistent and clearly defined with operational thresholds and equations.
Validation records include numerous historical replays showing strong model accuracy against past events like the European Gas Crisis and Fukushima.
The model's geographic and sectoral resolution is appropriate for its global stress-testing use cases.
The model aligns well with other CE models and accurately states scenario compatibilities, such as with the ECB BES and BoE CBES.
All data sources and methodologies are current, reflecting updates up to 2024 CE environment and scenario families.
WARN
Uncertainty treatment is only partially quantified; while limitations are acknowledged, specific uncertainty bounds for scenarios and outcomes are not fully defined.
CE USECE practitioners should use this model for stress testing scenarios dominated by policy fragmentation and sector fragility, serving as a downside boundary condition. It is particularly strong in identifying non-linear sector fragility and regulatory shock compression. However, avoid using for orderly transition analyses as it overstates pressure under these conditions. Be mindful of the adaptive dynamics and governance influences on fragility outcomes.
The model is grounded in IPCC AR6 and other authoritative sources, ensuring robust theoretical foundations.
Calibration sources such as UNEP Emissions Gap Report 2024 and IPCC AR6 WG3 are authoritative and appropriate.
The model architecture is consistent, with key mechanisms clearly defined and operationally detailed.
The geographic and sectoral resolutions are appropriate for a global, technology-focused model.
The data sources and methodologies are current, aligning with recent reports and technology assessments.
WARN
Uncertainty bounds are acknowledged but not comprehensively quantified, particularly in technology abatement estimates.
The validation record is not explicitly documented; there is no mention of historical data validation or comparable model benchmarks.
Integration with other CE models is noted as misaligned with NGFS policy scenarios, potentially limiting comprehensive analysis.
CE USECE practitioners should use this model to understand the scale of transformative climate solutions needed, especially focusing on technology gaps the model identifies. The model's strengths lie in its unique ability to quantify the 'breakthrough gap,' providing vital insights for technology investors and impact capital. However, caution should be exercised regarding the partially quantified uncertainties in technology deployment and the absence of explicit validation records, which might affect some decision-making scenarios.
CE Transition Opportunity Indexce-transition-opportunity
The model is based on well-established economic theories and principles, with all equations and mechanisms traceable to known literature and calibers.
The calibration sources listed, such as IEA and UNEP reports, are current and authoritative, with explicit acknowledgment of the strengths and limitations of each source.
The model includes a detailed record of validation against known historical events, which contributes to confidence in its forecasting ability.
Its geographic, temporal, and sectoral resolution is robust, providing insights at industry-sector, sub-sector, and company levels.
The model is well-integrated within the CE model family, with clear compatibility with other CE scenarios.
Data sources, scenario families, and methodologies are up-to-date, reflecting current market conditions and projections.
WARN
While key limitations and uncertainties are noted, the model could benefit from deeper quantifications of uncertainty bounds and possible scenario variations.
CE USEThis model should be employed by CE practitioners to identify high-opportunity sectors and companies during the energy transition. Its strengths lie in its detailed opportunity signaling, grounded in robust historical validation, and its integration with broader CE frameworks. However, practitioners should note that while uncertainties and limitations are acknowledged, a more quantified account of these factors could enhance decision-making confidence. The model is most effective when used alongside CE risk models to provide a balanced view of both opportunities and risks within the energy transition context.
CE Physical Hazard Cascade Modelce-physical-cascade
The model's theoretical foundation is well-grounded in compound event literature with published calibration sources.
The calibration uses authoritative sources like Swiss Re Sigma and IPCC AR6, ensuring high calibration integrity.
Structural completeness is demonstrated as the model's architecture aligns with described methodologies.
Validation records show consistent accuracy against historical case studies such as the 2017 California wildfires.
Resolution fitness is appropriate, offering global and region-specific insights for various sectors.
CE integration alignment is clear with precise scenario family support and integration with other CE models.
Operational currency is supported by up-to-date data sources such as CMIP6 and alignment with IPCC AR6 frameworks.
WARN
Uncertainty treatment could be more explicit in terms of bounds and sensitivity analysis, especially for new event types.
CE USECE practitioners should use the CE Physical Hazard Cascade Model to assess compound and cascading climate hazards in regions prone to multi-hazard risks, such as California and the Mediterranean. Its strengths lie in robust validation with comprehensive real-world case studies and integration with sector models for precision analysis. However, practitioners should be cautious of the model's limitations in handling novel event types and the extrapolation of parameters beyond observed ranges.
External Reference Scenarios
10NGFS, IEA, IPCC pathways used for CE calibration and alignment benchmarking
Scenario
Source
Type
Warming 2100
Checks (PI DP CA TR GR UC IG CU)
Findings
Status
Reviewed
NGFS Net Zero 2050ngfs-nz2050
NGFS Phase IV
Orderly Transition
1.4°C
✓✓✓!!!✓✓
3 found
Struct. Reviewed
2026-05-24
PASS
Pathway integrity: AR6 C1 ensemble-consistent; NZ2050 achieved across GCAM, REMIND-MAgPIE, and MESSAGE-GLOBIOM; carbon budget aligned with 1.5°C >50% IPCC AR6 WGIII C1 pathway; CDR scale within published feasibility bounds
Data provenance: NGFS Phase IV full technical documentation published; all three underlying IAMs peer-reviewed with published model documentation; AR6-aligned climate sensitivity used
Calibration alignment: the primary CE transition benchmark; carbon price path ($150-200/t in 2030, $500-700/t in 2050), emissions trajectory, and energy mix directly calibrate CE's policy and transition scenarios
CE integration gap: carbon price path and emissions trajectory directly importable into CE policy_instruments and financial_stress services; energy mix maps to grid_stability inputs; NGFS public data portal provides structured downloads
Currency: Phase IV (2023) is current; NGFS portal actively maintained with downloadable data files; technology costs post-IEA WEO 2023
WARN
Temporal resolution: 5-year timesteps only (2025, 2030, 2035...); annual granularity requires linear interpolation; CE scenario engine uses annual steps — interpolation error acceptable for smooth transition but document assumption
Geographic resolution: 10-11 global regions only; no country-level outputs for most variables; CE sub-national scenarios (Bangladesh, Gulf states, Vietnam) require downscaling using IEA WEO or World Bank country-level supplementation
Uncertainty coverage: no probability weights by design (central bank policy); within-scenario uncertainty limited to 3-IAM model spread (~10-15% on carbon prices); no formal Monte Carlo or probabilistic bounds — CE must apply external uncertainty ranges
CE USEPrimary CE transition benchmark. Use NGFS NZ2050 carbon price path and emissions trajectory as the calibration anchor for all CE net-zero scenarios. Interpolate 5-yr timesteps to annual. Downscale 10-region energy mix to country/sector using IEA WEO country data for CE sub-national scenarios. CE-aligned scenarios: gulf_wetbulb_survival, bangladesh_bay_of_bengal_transition, vietnam_jetp_transition.
NGFS Delayed Transitionngfs-delayed
NGFS Phase IV
Disorderly Transition
1.8°C
✓✓✓!!!✓✓
3 found
Struct. Reviewed
2026-05-24
PASS
Pathway integrity: delayed action before 2030 → abrupt carbon price shock 2030-2040 (2× NZ2050 path) → NZ by ~2060; 1.8°C budget consistent with scenario; asset stranding concentrated 2030-2040; all three IAMs internally consistent on the shock-then-recovery trajectory
Data provenance: same NGFS Phase IV full technical documentation and peer-reviewed IAM basis as NZ2050; no additional provenance concerns
Calibration alignment: primary CE delayed transition reference; carbon price spike ($600-800/t by 2040) is CE's calibration anchor for delayed transition risk in financial_stress and carbon_shock services
CE integration gap: carbon price shock path directly importable into CE financial_stress; NGFS Phase IV documentation includes country-level fossil fuel dependency exposure table usable as CE stranded asset exposure proxy
Currency: Phase IV (2023) current; NGFS portal actively maintained
WARN
Temporal resolution: same 5-year timestep limitation as NZ2050; annual data for the critical 2030-2035 shock window absent; linear interpolation in a non-linear shock regime misrepresents shock onset timing — particularly problematic for stranded asset valuation timing
Geographic resolution: same 10-11 regional limitation; stranded asset location impacts (fossil-heavy regions) not disaggregated to country level; requires IEA WEO country supplement for CE country-level mandate scenarios
Uncertainty coverage: single trajectory; no probability distribution on shock timing (2030 onset) or magnitude; no variants exploring earlier (2025) or later (2035) shock onset; CE must externally construct shock probability bounds
CE USECE primary delayed transition reference — use carbon price shock path ($600-800/t by 2040) as CE financial_stress and carbon_shock calibration anchor. Interpolate 5-yr timesteps with caution in the 2030-2035 window (shock onset non-linearity). Supplement with country-level fossil fuel dependency data for stranded asset exposure. CE-aligned scenarios: rust_belt_mandate, west_virginia_coal_to_clean, shanxi_dual_carbon_mandate.
NGFS Current Policiesngfs-current-policies
NGFS Phase IV
Hot House World
3.0°C+
✓✓!!!!!✓
5 found
Struct. Reviewed
2026-05-24
PASS
Pathway integrity: current emissions extrapolation consistent across all three IAMs; 3.0°C+ median consistent with current NDC track; energy mix reflects historical inertia without policy acceleration; consistent with IEA STEPS and IPCC AR6 WGIII BAU baseline
Data provenance: same NGFS Phase IV documentation and peer-reviewed IAM basis; no additional provenance concerns
Currency: Phase IV (2023) current; NGFS portal maintained
WARN
Calibration alignment: physical damage trajectory severely underestimated — NGFS IAMs use low-damage functions (~2-4% GDP at 3°C); CE uses 15-25% GDP impact at 3°C+ for physical risk; NGFS Current Policies economic loss estimates must NOT be cited in CE counterfactual_inaction narratives — use CE damage service outputs instead
Temporal resolution: same 5-year timestep limitation; annual data for near-term physical risk escalation absent
Geographic resolution: physical impacts at 3°C most geographically differentiated but remain at 10-11 regional level; country-level physical risk disaggregation absent
Uncertainty coverage: 3.0°C+ is a median; 4°C+ tail risk with tipping point dynamics not represented; no probability-weighted physical outcomes
CE integration gap: emissions baseline and carbon price floor (near-zero) safe to import; physical damage, GDP impact, and economic cost outputs must NOT be imported without explicit override — NGFS IAMs understate physical risk at T>2.5°C by 3-7×; CE damage service must override
CE USEUse NGFS Current Policies as the NO-ACTION emissions baseline and carbon price floor reference only. Do NOT use NGFS Current Policies physical damage or economic cost estimates — IAMs understate physical risk at T>2.5°C by 3-7×. CE counterfactual_inaction narratives must use CE damage service outputs, not NGFS economic outputs. Primary CE use: counterfactual baseline for gap analysis (NZ2050 policy benefit vs. current policies harm) and emissions trajectory reference for CE BAU scenarios.
IEA Net Zero Emissions by 2050iea-nze-2050
IEA WEO 2025
Orderly Transition
1.5°C
✓✓✓!!!✓✓
3 found
Struct. Reviewed
2026-05-24
PASS
PI: Internal carbon budget consistent with 1.5°C; technology deployment pathway coherent and cross-validated against IPCC AR6 SSP1-1.9 envelope
DP: IEA WEO 2025 with full sector-model methodology documentation; widely cited and peer-benchmarked across IAM community
CA: Technology roadmaps (electricity capacity additions, EV deployment, industrial decarbonisation rates) directly usable as CE fleet and sector calibration inputs; more granular than NGFS for technology vectors
IG: Carbon price trajectories and tech deployment rates importable for grid_stability, energy_system_opt, and emissions overlays with minimal bridging; no physical damage conflict
CU: WEO 2025 edition; AR6-aligned carbon budgets; current as of May 2026
WARN
TR: 5-year intervals only (2025–2050); annual interpolation required for CE time-step modeling — use linear interpolation or smooth spline between IEA waypoints
GR: Regional aggregates (Americas, Europe, Asia Pacific, Africa, Middle East); country-level only for major economies — sub-national CE scenarios require manual disaggregation
UC: Single deterministic trajectory; no probability envelope or sensitivity variants published; CE must construct own uncertainty range around IEA central path (suggest ±15% on deployment rates)
CE USEPrimary CE technology benchmark for 1.5°C orderly transition. Carbon price and technology deployment rates are the most granular available for CE fleet capacity and grid_stability calibration. Use alongside NGFS NZ2050 for cross-validation: IEA NZE provides better technology detail; NGFS NZ2050 provides better financial sector impact data. Not a substitute for NGFS on transition risk financials.
IEA Announced Pledges Scenarioiea-aps
IEA WEO 2025
Moderate Transition
1.7°C
✓✓!!!!✓✓
4 found
Struct. Reviewed
2026-05-24
PASS
PI: Pathway consistent with full NDC + net-zero pledge compliance; 1.7°C carbon budget correctly bounded and internally consistent
DP: IEA WEO 2025 methodology; country-level policy tracking basis documented with enacted-law vs. pledge distinction
IG: Country-level sectoral capacity additions and policy trajectories directly usable for CE policy_overlay calibration against stated national targets
CU: WEO 2025 edition; current as of May 2026
WARN
CA: APS assumes 100% NDC compliance — historically, implementation rates run 60–80% of pledged ambition; CE scenarios using APS as a baseline should apply a credibility discount (20–30% policy slippage) or treat APS as the optimistic bound of a STEPS–APS range
TR: 5-year intervals only; annual interpolation required for CE time-step modeling
GR: Country-level only for major economies (US, EU, China, India); limited granularity for emerging markets and no sub-national resolution for CE local contexts
UC: Single deterministic trajectory; IEA publishes no probabilistic NDC compliance variants; policy slippage uncertainty not quantified
CE USEReference scenario for CE moderate transition cases — governments do what they say, but no more. Best positioned as the upper bound of a STEPS–APS policy range (STEPS = enacted floor, APS = pledge ceiling, NZE = aspirational target). Apply 20–30% credibility discount to policy_overlay parameters when using APS country trajectories as direct CE calibration inputs.
IEA Stated Policies Scenarioiea-steps
IEA WEO 2025
Insufficient Transition
2.4°C
✓✓!!!!!✓
5 found
Struct. Reviewed
2026-05-24
PASS
PI: Internally consistent with enacted-policy trajectory; 2.4°C outcome correctly follows from current policy set; no internal pathway contradictions
DP: IEA WEO 2025 methodology; enacted-law audit documented at country level with clear boundary between enacted and announced policies
CU: WEO 2025 edition; current as of May 2026
WARN
CA: STEPS energy transition outputs (capacity additions, fuel mix, demand by sector) are directly usable for CE BAU fleet modeling — but IEA does not model physical climate impacts. Do NOT import STEPS as a physical damage baseline; bridge to IPCC SSP2-4.5 or SSP3-7.0 for physical risk at equivalent warming levels
TR: 5-year intervals only; annual interpolation required for CE time-step modeling
GR: Regional and major-economy level; limited sub-national resolution for CE local scenario contexts
UC: Single deterministic trajectory; 2.4°C outcome carries wide physical risk uncertainty not captured in STEPS outputs — tail risk at 2.4°C is significant and unstated
IG: No carbon price trajectory (policies-only, no market signal); CE must supply its own carbon price overlay independently. No physical damage pathway — users must manually bridge to IPCC SSP2-4.5 for climate physical impacts at this warming level. BAU fleet additions and fuel-type capacity are the only directly importable outputs.
CE USECE BAU reference: use STEPS fleet capacity additions and fuel-mix trajectories as the no-transition baseline for CE scenario comparisons. Do NOT use as a physical damage reference — pair with IPCC SSP2-4.5 (or SSP3-7.0 for tail risk) for physical climate impacts at equivalent warming levels. Carbon price = zero in STEPS; CE policy_overlay must be added independently for any transition cost analysis.
SSP1-1.9ipcc-ssp1-19
IPCC AR6
Aggressive Mitigation
1.0–1.8°C
✓✓✓!!✓✓✓
2 found
Struct. Reviewed
2026-05-24
PASS
PI: C1 category multi-model ensemble; carbon budget, energy pathway, and socioeconomic narrative internally consistent; net-negative CO₂ by 2060s explicitly modelled
CA: Physical climate outputs (temperature anomaly, precipitation, extremes, SLR) directly usable as CE climate override calibration inputs; CE mandate achievement rates benchmarkable against SSP1-1.9 technology and policy trajectories
UC: Multi-model ensemble with 5th–95th percentile ranges published; superior uncertainty quantification relative to IEA or NGFS single-trajectory scenarios
IG: Physical risk parameters (heat_stress, flood_risk, water_stress) available from AR6 WGI Atlas regional downscaling; maps directly to CE climate_override layer; use as "avoided damage" counterfactual in CE best-case mandate achievement framing
CU: IPCC AR6 (2021–2022); AR7 not yet published; current
WARN
TR: Physical climate projections provided as decadal means from GCMs (2021–2040, 2041–2060, etc.); IAM energy/emissions data available annually in IAMC database but climate variables require temporal interpolation for CE annual time-step modeling
GR: IAM socioeconomic and energy outputs are regional aggregates (R5/R10 regions); WGI Atlas provides country and sub-national climate downscaling but at coarser resolution than CE local scenarios require — sub-national CE contexts need additional downscaling step
CE USECE best-case physical risk baseline and avoided-damage counterfactual. SSP1-1.9 physical parameters define the lower bound of CE climate impact ranges for scenarios that achieve aggressive decarbonisation. Note: net-negative emissions assumption (heavy CDR/BECCS post-2050) should be flagged explicitly in any CE scenario that draws on SSP1-1.9 for long-run physical damage avoidance — CDR deployment at that scale is not a CE default assumption.
SSP2-4.5ipcc-ssp2-45
IPCC AR6
Intermediate
2.1–3.5°C
✓✓✓!!✓✓✓
2 found
Struct. Reviewed
2026-05-24
PASS
PI: Internally consistent; 2.1–3.5°C multi-model spread reflects SSP2 "middle of the road" socioeconomic trajectory; no internal contradictions
DP: IPCC AR6 WGIII IAMC database; most widely calibrated and cited IPCC scenario in climate-economics literature
CA: Highest CE calibration compatibility of any IPCC scenario — economic damage estimates at 2–3°C are well-constrained; CE heat_stress, flood_risk, and water_stress parameters have the richest literature support against SSP2-4.5 regional downscaling
UC: Multi-model ensemble with explicit 5th–95th percentile ranges; 2.1–3.5°C spread publicly documented; adequate for CE uncertainty banding
IG: Primary CE physical risk calibration scenario — maps directly to CE climate_override layer for insufficient-action narratives; pairs naturally with IEA STEPS as the energy-transition complement at equivalent warming
CU: IPCC AR6 (2021–2022); current
WARN
TR: Physical climate projections as decadal means from GCMs; annual interpolation required for CE time-step physical risk modeling
GR: IAM outputs at R5/R10 regional aggregates; WGI Atlas provides country downscaling for climate variables but sub-national CE scenarios require additional spatial disaggregation
CE USEDefault CE physical risk calibration scenario for insufficient-action narratives (2–3°C range). Use SSP2-4.5 as the primary source for heat_stress, flood_risk, water_stress, and agricultural loss parameters in CE scenarios where current policies are roughly maintained. Pairs with IEA STEPS as the energy-transition complement — STEPS provides the BAU fleet trajectory, SSP2-4.5 provides the physical damage pathway at that level of ambition.
CA: Physical damage functions become highly nonlinear above 3°C — SSP3-7.0 central damage estimates carry significantly wider model-to-model spread than SSP2-4.5; CE damage_overlay parameters sourced from this scenario should include explicit uncertainty multipliers (suggest 1.5×–2.5× central estimates for T>3°C)
TR: Physical climate projections as decadal means; annual interpolation required for CE time-step physical risk modeling
GR: IAM outputs at R5/R10 aggregates; physical climate downscaling available at country level but sub-national resolution requires additional step
UC: Published ensemble spread likely understates true physical risk at T>3°C — tipping point risks (Greenland ice sheet destabilisation, AMOC slowdown, permafrost carbon release) are not fully captured in central CMIP6 ensemble; CE tail-risk scenarios using SSP3-7.0 should explicitly flag tipping point exclusion
IG: CE counterfactual_inaction physical risk costs reference SSP3-7.0 damage estimates, but at 3–4°C+, CE services (water_stress, land_valuation, agricultural_loss) encounter nonlinearities requiring explicit tipping-point overlays not provided by IPCC; CE users must manually add tipping-point risk premium on top of SSP3-7.0 central estimates for high-end scenario framing
CE USECE counterfactual_inaction physical risk reference and tail-risk framing scenario. Use SSP3-7.0 as the high-end physical damage baseline for CE scenarios analysing inaction costs and physical risk under fragmented policy response. Critical caveat: published central damage estimates understate true tail risk at T>3°C — always apply an explicit tipping-point risk premium in CE analyses using this scenario. Pairs with NGFS Current Policies (transition) for a full inaction scenario stack.
SSP5-8.5ipcc-ssp5-85
IPCC AR6
Very High Emissions
3.3–5.7°C
✓✓!!!!!✓
5 found
Struct. Reviewed
2026-05-24
PASS
PI: Internally consistent; SSP5 fossil-fuelled development narrative drives 3.3–5.7°C range; CMIP6 ensemble coherent with emissions pathway
DP: IPCC AR6 WGIII IAMC database; CMIP6 physical projections peer-reviewed; consistent with previous AR5 RCP8.5 literature body
CU: IPCC AR6 (2021–2022); current
WARN
CA: SSP5-8.5 is explicitly characterised by IPCC AR6 as a "low likelihood, high impact" scenario — it requires active expansion of fossil fuel use inconsistent with current energy economics. CE users who present SSP5-8.5 as a business-as-usual or central projection are misrepresenting AR6 consensus. Use ONLY as upper-bound tail risk; pair with explicit framing statement
TR: Physical climate projections as decadal means from GCMs; annual interpolation required for CE time-step physical risk modeling
GR: IAM outputs at R5/R10 regional aggregates; WGI Atlas provides country downscaling for climate variables only
UC: Published CMIP6 ensemble spread likely understates true tail risk above 4°C — tipping point cascades (WAIS destabilisation, permafrost carbon feedback, Amazon dieback) are not fully represented in central ensemble projections; physical risk in the upper tail is wider than published percentile bands suggest
IG: CE tail risk parameters (flood_risk, heat_stress, wet-bulb, cyclone intensity) importable from AR6 WGI Atlas at SSP5-8.5 forcing level. At >4°C, CE damage functions encounter deep nonlinearity — tipping-point risk overlays required. Explicitly label any CE output using SSP5-8.5 physical parameters as tail risk / stress test, not median projection.
CE USECE upper-bound physical risk stress test and tail risk framing only. Use SSP5-8.5 physical parameters (heat_stress, flood_risk, wet-bulb exceedance, SLR) as the ceiling for CE tail risk scenario banding. MANDATORY framing requirement: any CE analysis citing SSP5-8.5 must include the IPCC AR6 "low likelihood, high impact" designation — it is not a BAU scenario. Pairs with NGFS Current Policies (transition narrative) for the worst-case combined physical + transition scenario stack.
Reviews stored in data/reviews/reviews.json.
Pull scenario JSON for review via /api/scenarios.