Radiative forcing–based accounting

What a tonne of gas or pollutant is worth depends on when it is cut.

Carbon markets convert GHGs to CO₂-equivalents with one fixed number, GWP₁₀₀. This paper proposes crediting it from the underlying physical curve instead: the same IPCC physics, integrated over the time that actually remains in the policy window. A short-lived climate pollutant like methane, for example, abated early counts for more, because more of its benefit arrives in time to matter.

Shivang Agarwal, Saloni Srivastava, Bhhavya Kapoor, Zerin Osho and Daniele Visioni (2026), “Radiative forcing-based accounting (RFA): a dynamic framework to replace static GWPs in carbon markets,” Environmental Research Letters 21 064024.

Fig. 1 · The RFA multiplier M(T): tonnes of CO₂-equivalent per tonne of methane, as a function of the horizon. The familiar GWP₂₀ and GWP₁₀₀ are two points on this curve. Computed from IPCC AR6 impulse-response functions, eqs. (1)–(5) of the paper.

RFA Credit Calculator (Methane and CO₂)

The same equations, applied to a project of your own

RFA is fully computable from project basics. Set an abatement schedule and a policy horizon; the per-year multipliers and the gap against static crediting follow directly from the equations.

Fig. 2 · Calculator

What would RFA mean for a project like yours?

Enter an abatement schedule and a policy horizon. Every figure recomputes from the paper's equations: the same physics, applied to your numbers. Avoided CO₂ counts one-to-one under every method (equation 7); the difference between methods is entirely about methane. (Input negative values to account for emission leakage.)

RFA credits (dynamic)
2.45M tCO₂e
Year-1 CH₄ multiplier 30.0 → final 27.9 · includes 1.00M t CO₂ at 1:1
Static GWP₁₀₀ SAR (21)
2.05M tCO₂e
What CDM-era rules would credit
RFA uplift vs SAR
+19.4%
vs AR6 static (27.9): +2.2%
RFA (dynamic)Static GWP SAR (21)
View as table
Calculator results by year
YeartCH₄CH₄ multiplierRFA tCO₂e (incl. CO₂)GWP SAR tCO₂e (incl. CO₂)
Yr 15,00030.0250,136205,000
Yr 25,00029.8248,865205,000
Yr 35,00029.5247,617205,000
Yr 45,00029.3246,393205,000
Yr 55,00029.0245,191205,000
Yr 65,00028.8244,012205,000
Yr 75,00028.6242,854205,000
Yr 85,00028.3241,717205,000
Yr 95,00028.1240,600205,000
Yr 105,00027.9239,503205,000
Fig. 3 · Credit share by gas

Where the credit comes from, as the horizon grows

For the mix entered above, methane dominates the credit at short horizons and fades as the window lengthens, because its forcing is concentrated up front while CO₂ persists. This is the whole time-horizon debate in one picture.

Share from CH₄Share from CO₂
View as table
Credit share by gas at selected horizons
Horizon TShare from CH₄Share from CO₂
1 yr85.4%14.6%
5 yr84.8%15.2%
10 yr83.6%16.4%
20 yr80.4%19.6%
50 yr70.3%29.7%
100 yr58.2%41.8%

Background

Where the frozen number came from

Four short notes for readers new to carbon markets. If you already know what GWP₁₀₀ is and why its horizon is contested, skip to the figures.

  1. Carbon credits

    Projects that prevent greenhouse-gas emissions earn credits, each standing for one tonne of CO₂ kept out of the atmosphere, and the credits are bought and sold on carbon markets. The system needs an exchange rate the moment the avoided gas is anything other than CO₂.

  2. Short-lived gases (like methane) warm on different clocks in comparison to CO₂

    For example, a tonne of methane heats the planet far more than a tonne of CO₂, but it fades from the atmosphere in about 12 years, while CO₂ lingers for centuries. Any exchange rate between the two therefore depends on how long a period you score them over.

  3. Markets settled on one frozen number

    The rate in use is GWP₁₀₀: a gas’ warming impact averaged over 100 years, expressed in tonnes of CO₂. Under this rate, a gas cut today and cut in a decade are rewarded identically; the timing of the benefit is invisible.

  4. The paper lets the rate move

    Radiative Forcing-based Accounting keeps the same IPCC physics but integrates it over the time remaining in the policy window, so a gas cut early or later in a project counts differently. By the final project year the rate settles exactly on the current official value: nothing is invented, and the two methods agree at the boundary.

Why it matters Cutting short-lived gases like methane is the fastest available lever for slowing warming within the next few decades, yet the projects doing that work are paid at an exchange rate frozen in 1995. When the paper re-scores four real projects with the physical curve, the frozen rate turns out to undervalue them by 36 to 40 percent. Correcting this requires no new equipment and no new paperwork; registries would swap one hard-coded constant for a small calculation.

The physics

Every static GWP is one point on a single curve

Methane traps heat intensely but decays within decades; CO₂ lingers for centuries. Integrating their radiative forcing over any horizon T gives the equivalence multiplier for that horizon. RFA assigns each project a multiplier appropriate to their lifetime.

Fig. 4 · Interactive

The multiplier is a curve, not a constant.

Every static GWP is one point on this curve. RFA credits each project year at the point the policy horizon actually implies (T = H − L + t), so early years of a project earn more per tonne than late years.

RFA multiplier M(T)Your project's crediting yearsGWP₁₀₀ AR6 (27.9)GWP₂₀ AR6 (81.2)GWP₁₀₀ SAR (21) · GWP₂₀ SAR (56)

With H = 100 and a 10-year project, year 1 is credited at 30.0 tCO₂/tCH₄ (T = 91) declining to 27.9 in year 10 (T = 100), converging to the fixed GWP value at the boundary, in alignment with the study.

View as table
RFA multiplier by project year
Project yearEffective horizon TMultiplier (tCO₂/tCH₄ = tCO₂e)
19130.0
29229.8
39329.5
49429.3
59529.0
69628.8
79728.6
89828.3
99928.1
1010027.9

The evidence

Four CDM projects, re-credited

The paper applies RFA to four methane-abatement projects registered under the UNFCCC Clean Development Mechanism: landfill gas, municipal waste, livestock manure, and industrial wastewater. CDM rules credited all four at a flat GWP₁₀₀ of 21, a value fixed in 1995.

Fig. 5 · Case study

Annual CO₂e credits: Xingfeng Landfill, 20082014

Captured and destroyed landfill methane while generating electricity. The largest of the four projects by an order of magnitude.

RFA (dynamic)Static GWP₁₀₀ SAR (21), used by CDM rulesStatic GWP₁₀₀ AR6 (27.9)

Annual profiles are digitized from the paper's supplementary figures S1 to S4; totals are exact and cumulative credits match supplementary Table S1 to three significant figures. Per-year multipliers match Figure 2 of the paper.

View as table
Annual credits for Xingfeng Landfill
YeartCH₄ avoidedRFA multiplierRFA tCO₂eGWP SAR tCO₂eGWP AR6 tCO₂e
200839,21829.31,148,245823,5781,094,183
200944,20929.01,283,762928,3971,233,442
201047,53728.81,369,176998,2771,326,282
201147,53728.61,358,167998,2771,326,282
201244,44728.31,259,779933,3891,240,074
201341,59528.11,169,648873,4921,160,497
201438,74327.91,080,943813,5961,080,920
Methane avoided (published total)
303,286 t
Landfill gas recovery & electricity generation
RFA credits · H = 100
8.67M tCO₂e
vs static GWP₁₀₀ (SAR 21)
+36.1%
Paper reports +36%: the credits GWP leaves unrecognized
vs GWP₁₀₀ (AR6 27.9)
+2.5%
Against updated AR6 science, RFA and static values converge at the boundary
Fig. 6 · Heatmap

Multiplier by project year: every project, every year

The paper's Figure 2, recomputed live. Bold: RFA multiplier (tCO₂/tCH₄ = tCO₂e). Small: Δ vs the static SAR value (21) that CDM rules actually applied. Hatched cells fall outside the project's crediting period.

Year 1Year 2Year 3Year 4Year 5Year 6Year 7Year 8Year 9Year 10Year 11
Xingfeng Landfill29.38.329.08.028.87.828.67.628.37.328.17.127.96.9
Chandigarh MSW30.09.029.88.829.58.529.38.329.08.028.87.828.67.628.37.328.17.127.96.9
Brazil Livestock30.09.029.88.829.58.529.38.329.08.028.87.828.67.628.37.328.17.127.96.9
Tamil Nadu Wastewater30.39.330.09.029.88.829.58.529.38.329.08.028.87.828.67.628.37.328.17.127.96.9

Longer projects start with higher multipliers (a longer remaining window), and every row converges to the AR6 boundary value of 27.9 at T = 100 by its final year.

View as table
RFA multipliers by project and year
ProjectY1Y2Y3Y4Y5Y6Y7Y8Y9Y10Y11
Xingfeng Landfill29.329.028.828.628.328.127.9
Chandigarh MSW30.029.829.529.329.028.828.628.328.127.9
Brazil Livestock30.029.829.529.329.028.828.628.328.127.9
Tamil Nadu Wastewater30.330.029.829.529.329.028.828.628.328.127.9
+36 to 40%
Credits Undercounted by GWP₁₀₀
Across four UNFCCC CDM methane project case studies, RFA recognizes 36 to 40 percent more climate benefit than the static rate they were paid under (57 to 68 percent at a 20-year horizon).
117 → 28
The multiplier's range, in tCO₂/tCH₄ = tCO₂e
From a 1-year to a 100-year integration window. Every static GWP ever used is one point in this range.
None
New monitoring required
Measurement, reporting, and verification are unchanged. One hard-and-fast constant is replaced with one dynamic value during GHG accounting by registries, project developers, regulators, and all associated market participants.

The method

The four equations behind this page

RFA rests on the same physical science as GWP itself: IPCC AR6 radiative efficiencies and impulse-response functions. What changes is how time is aggregated. The policy horizon becomes an explicit, governable parameter rather than a constant hidden inside a table.

Step 1

Atmospheric decay (impulse-response functions)

CO₂ persists across four sink timescales (Joos et al. 2013, IPCC AR6); CH₄ decays with a single 12.4-year lifetime.

IRFCO₂(t) = 0.217e−t/10⁵ + 0.224e−t/394 + 0.282e−t/36.3 + 0.276e−t/4.3
IRFCH₄(t) = e−t/12.4
Step 2

Integrated forcing (AGWP)

Radiative efficiency × persistence, integrated over the evaluation horizon T. AR6 values: RE꜀ₒ₂ = 1.33×10⁻⁵, RE꜀ₕ₄ = 5.70×10⁻⁴ W m⁻² ppb⁻¹ (indirect-inclusive).

AGWPg(T) = 0T REg · IRFg(t) dt
AGWPCH₄(T) = RECH₄ · τ (1 − e−T/τ)
Step 3

The dynamic multiplier

The AGWP ratio, converted to mass terms. A continuous function of T rather than a fixed constant.

MRFA(T) =AGWPCH₄(T)AGWPCO₂(T)·44.0116.04
  • M(1) = 117.1 · M(20) ≈ 81–82 (GWP₂₀ AR6) · M(100) = 27.9 (GWP₁₀₀ AR6)
  • Static GWPs are single points on this curve. RFA keeps the whole curve.
Step 4

The dynamic horizon & crediting

Policymakers declare one policy horizon H, an explicit and governable choice. Each project year t of an L-year project is then integrated over the window that actually remains.

Hi = H − L + t
Credits(t) = ECH₄(t) × MRFA(Hi)
  • No new MRV: registries swap a fixed constant for a computed multiplier; nothing else changes.
  • Governance: H must be standardized, project length L capped, fungibility and banking rules defined.
Shivang Agarwal, Saloni Srivastava, Bhhavya Kapoor, Zerin Osho and Daniele Visioni (2026)
Radiative forcing-based accounting (RFA): a dynamic framework to replace static GWPs in carbon markets. Environ. Res. Lett. 21 064024 · CC BY 4.0