Climate : Scientific Certainties

- CLIMATE -
What is Established with Scientific Certainty
and How

Measurements, physical laws, and observations
independent of numerical models and future projections

By the Scientific Committee "Terre & Climat" (TM) - France, 5 February 2026, revised 23 February 2026, v.1.7

INTRODUCTION

Contemporary climate debate often suffers from confusion between observed facts, established physical laws, results derived from numerical modeling, and future projections. Yet a substantial part of climate science rests on direct measurements and fundamental physical laws, independent of any socio-economic scenario.

This document presents what is considered scientifically established, in the strict sense of the term.

1. The Atmospheric Greenhouse Effect Is Demonstrated and Measured

The outgoing infrared radiation from the Earth to space, measured by satellites since the 1970s, is currently about 238–239 W/m² of irradiance.

Planck’s law describes the spectral radiance of a black body, providing the complete and continuous spectral distribution over all wavelengths.
The Stefan–Boltzmann law (M = εσT⁴) is the integral of Planck’s law over all wavelengths and allows temperature to be inferred from the received irradiance M. For atmospheric gases, which do not exhibit a continuous spectrum, the Stefan–Boltzmann law is not directly applicable. Their spectral emissivity ε(λ) varies strongly with wavelength: some wavelengths radiate very weakly (atmospheric windows), while others behave as quasi-black bodies (ε ≈ 1). Only an apparent radiative temperature, dependent on the measured spectral domain, can be defined indirectly.

The outgoing flux to space — the only one that allows Earth to maintain thermal equilibrium — corresponds to an equivalent radiative temperature of approximately –18 °C (which is still high compared with absolute zero). Its spectrum is strongly shaped by the absorption bands of atmospheric greenhouse gases (H₂O, CO₂, etc.) and no longer corresponds to the continuous black-body spectrum emitted by the Earth’s surface.

This is a measured fact, not a theoretical construction.

By contrast, the measured mean surface temperature is approximately +15 °C. This apparent difference of about 33°C necessarily implies an atmosphere partially opaque to infrared radiation and demonstrates the existence of a global atmospheric greenhouse effect, resulting from infrared absorption and re-emission by atmospheric gases.

Spectral observations show that the infrared radiation escaping to space (apparent mean temperature ≈ –18 °C) originates from a global equivalent mean altitude of roughly 4–6 km, rather than primarily from the surface. This result, established by the work of Goody and Yung (1952–1989) and subsequently confirmed by modern satellite measurements, constitutes a direct observational demonstration of the atmospheric greenhouse effect.

The probability coefficients defined by Einstein play a role in complex radiative phenomena within the atmosphere, notably the absorption and spontaneous and induced emission of infrared radiation by greenhouse gases. Spontaneous emission (coefficient Aul) and induced emission (coefficient Bul) coexist and are proportional to the wavelength ν³. Aul = (8πhν³/c³) × Bul,
where h is Planck's constant and c is the speed of light.
At the principal wavelength of CO₂, isotropic spontaneous emission is 27 times more intense than induced emission.

A simple illustration: if a warm source (the Earth’s surface) is covered by even a thin curtain, the source must become warmer in order to heat the region above the curtain (toward space) to the same extent.

2. Global Mean Temperature Has Increased for More Than a Century

Temperature series from independent datasets (land stations, ocean buoys, satellites) show a clear increase in global mean temperature since the late 19th century: approximately +1.2°C.

This trend is observed outside periodic El Niño events and is confirmed by multiple independent measurement methods.

IPCC AR6 reference : WG1, Chapter 2

Global surface temperature evolution since 1880
(According to NASA/GISS)

Lower tropospheric temperature measured by satellites
(According to UAH v6.1) For 47 years

In France, located in the Northern Hemisphere, which is warming faster due to a smaller oceanic surface area, Météo-France reports a warming of approximately +2.2°C since the 19th century.

3. Anthropogenic Increase in Atmospheric CO₂

Continuous measurements since 1958 show an increase in atmospheric CO₂ from about 315 ppm to more than 420 ppm (+33%). Isotopic analyses of carbon (¹²C, ¹³C, ¹⁴C) indicate that this increase — partially limited by oceanic and biospheric uptake — originates primarily from fossil-fuel combustion and deforestation. This attribution is robustly established.
Relative to earlier pre-industrial levels (~280 ppm), the increase amounts to approximately +50%.

Methane and nitrous oxide concentrations, of lesser importance for the total
greenhouse effect, have also increased significantly since the pre-industrial era,
and their anthropogenic origin is well documented.

4. Variations in Solar Irradiance Are Small Over the Recent Period

The Baseline Surface Radiation Network (NASA-BSRN) provides quasi-continuous, long-term, in-situ measurements of broadband surface irradiance (solar and thermal infrared radiation), along with associated parameters, through a global network of more than 70 sites.

Satellite measurements of Total Solar Irradiance (TSI) since 1978 show solar cycles (notably the well-known ~11-year cycle) with no increasing trend over recent decades (TSI ≈ 1361 W/m²).

The IPCC AR6 : WG1, Chapitre 2 (Changes in Climate System Drivers - Solar and Orbital Forcing) concludes that the recent solar contribution to radiative forcing is very small compared with that of greenhouse gases, with no recent upward trend.

Several recent studies even indicate a slight weakening of solar irradiance over recent decades.

   Other thermal influences — negligible energies % relative to the Sun
Geothermal energy (including volcanism)                      0,04 %
   Human activities (industry, combustion, etc.)                0,02 %
   Tidal energy                                                                         0,003 %
   Heat flux from Earth’s core                                               0,04 %

5. Atmospheric Infrared Back Radiation Is Measured

SURFACE ENERGY BALANCE W/m2

Downwelling longwave radiation (DLR), often referred to as infrared atmospheric back radiation, has been directly measured for several decades using calibrated pyrgeometers. Observations from the Baseline Surface Radiation Network Baseline Surface Radiation Network (BSRN) show a global mean irradiance on the order of 340–345 W/m² - (344 ± 3 W/m² according to a recent synthesis article).
This irradiance varies strongly with latitude; values of about 70 W/m² are observed near the poles.

In FRANCE, downwelling longwave irradiance is measured at SIRTA (Palaiseau).

These measurements are purely observational: they do not depend on climate models or on any disputable theoretical assumption.

Due to back radiation, the Earth's surface, heated to 15°C and which, according to the Stefan-Boltzmann law, should radiate approximately 297 W/m² depending on its temperature, has a "net" radiation (cooling the surface) of only about 53 W/m² (= 397 - 344). This is generally poorly understood.

IPCC AR6 reference : WG1, Chapitre 7 (Radiative Forcing).

It is nevertheless important to note that the Earth’s surface warms through global energy
re-equilibration following changes induced by radiative forcings, rather than through
the direct reception at the surface of individual forcings considered in isolation.

6. Clouds Emit Infrared Radiation Toward the Surface

Clouds, composed of liquid water droplets or ice crystals at temperatures far above absolute zero, emit infrared radiation in accordance with Planck's law, according to their spectral properties, both upward and downward. This emission is spectrally measured and routinely observed (including the well-known phenomenon of warmer cloudy nights).
Low-level clouds (stratus, stratocumulus) can increase downwelling longwave radiation by 50 to 100 W/m².
The IPCC states that the infrared effect of clouds is well established, even though their net global radiative balance — due to their high variability and their impact on albedo — remains a major source of uncertainty.

IPCC AR6 reference : WG1, Chapitre 7

7. Spectral Line Broadening: Undisputed Physics

Atmospheric gases, particularly CO₂, absorb infrared radiation in specific spectral lines that broaden and strengthen (line wings) through fundamental physical mechanisms : pressure broadening, molecular collisions, Doppler effects, and quantum resonance phenomena (notably Fermi resonance for CO₂).
This explains why increases in atmospheric CO₂ continue to have a measurable effect, although increasingly attenuated, even though the main absorption band near 15 µm is already close to saturation.

CO₂ cannot become globally “saturated,” because the effect of increasing concentration follows a logarithmic law : ln(CO2_n/CO2_0) and therefore never becomes zero. Moreover, as CO₂ concentration increases, the effective altitude of infrared re-emission shifts upward to colder layers, requiring a slight warming to re-establish radiative balance with space. These mechanisms are measured in laboratory experiments and incorporated into reference spectroscopic databases such as HITRAN.

IPCC AR6 reference: WG1, Chapter 6, pp. 853–854 (Radiative processes)

8. Sea level

Coastal tide gauges indicate a rise in global mean sea level of approximately 18–22 cm since 1900, corresponding to an average rate of 1.2–1.7 mm/year during the 20th century, corrected for vertical land motion (isostatic adjustment).

At Brest (France), where tide gauges have recorded sea level for three centuries — a unique dataset worldwide — observations show an increase of about 30 cm over the past 300 years.

Since the beginning of satellite altimetry in 1993, measurements show a mean rise of approximately 3.2–3.5 mm/year, with a recent acceleration reaching 3.5–4.0 mm/year(TOPEX/Poseidon, Jason-1/2/3).

This rise results primarily from thermal expansion of seawater and melting of continental ice. It is a directly measured phenomenon, independent of climate models.
These contributions are observed using satellite gravimetry (GRACE / GRACE-FO), altimetry, and ocean measurements (ARGO).

9. IPCC Temperature Projections Are High and Imprecise

The temperature projections published by the IPCC are primarily based on numerical climate models used to explore a range of possible future greenhouse gas emission scenarios.
These projections do not constitute deterministic forecasts, but rather conditional simulations, dependent both on the socio-economic assumptions adopted and on internal model parameters, some of which—particularly climate sensitivity and cloud feedbacks—remain imperfectly constrained by observations.

As such, the range of projected warming should be interpreted with caution. The discrepancies observed between simulated trajectories and the temperature evolution actually observed over recent decades indicate that climate models tend to produce a wide dispersion of outcomes, including high-warming scenarios whose real-world probability remains a matter of scientific debate.
These projections therefore serve as tools for exploring potential climate risks, rather than as quantitatively certain descriptions of the future evolution of the climate system.

Global mean temperature (IPCC page n°87) Sea Surface Temperature (Dr Roy Spencer)

CONCLUSION

The phenomena presented in this document are based on direct instrumental observations and fundamental physical laws. They constitute a robust scientific foundation, independent of numerical models, assumptions, or prospective scenarios.

Current scientific debates concern complex feedbacks and the future magnitude of warming, not these established foundations.

UNCERTAINTIES

What Climate Science Cannot Yet Assert

Climate science today rests on a robust foundation of observed facts and established physical laws. Nevertheless, despite abundant observations and solid underlying physics, several central questions remain open. Significant uncertainties persist regarding feedbacks, regional scales, and the precise future evolution of the climate system. It is essential to distinguish these uncertainties clearly from established facts.

Acknowledging these limits does not weaken science; on the contrary, it is a fundamental methodological requirement.

A. The Precise Long-Term Value of Climate Sensitivity λ to CO₂

Climate sensitivity (the global temperature increase associated with a doubling of CO₂) is not directly measured. Estimates rely on paleoclimate reconstructions, recent observations, and climate models. Despite decades of research, the uncertainty range remains wide. IPCC AR6 explicitly acknowledges that this value is not precisely known and remains one of the most influential and uncertain parameters in future projections (additional warming beyond the minimal, incompressible spectroscopic Planck forcing).

B. The Exact Role of Cloud Feedbacks

Clouds influence both reflected solar radiation (cooling effect) and infrared radiation emitted toward the surface (warming effect). While their downwelling infrared effect is measured, their net global feedback (positive or negative) remains poorly constrained. Differences in cloud representation account for a significant fraction of the spread among climate model results.

C. The Amplitude of Internal Natural Variability on Decadal Scales

Oceanic and atmospheric systems exhibit natural oscillations (ENSO, PDO, AMO, etc.) whose true amplitude, duration, and interaction with long-term trends are not fully quantified. It is therefore not possible to precisely attribute, year by year or decade by decade, the respective contributions of external forcings and internal variability.

D. Direct Transposition of Global Trends to Regional Scales

Trends observed at the global scale do not translate mechanically to regional or local scales. Regional projections depend strongly on models, show marked divergences, and involve uncertainties greater than those of global means.

E. The Exact Future Evolution of the Climate System

Climate projections depend on emission scenarios, socio-economic assumptions, and internal model parameters. They are not predictions in the strict sense, but conditional explorations of possible futures. Science therefore cannot assert a single, certain climate trajectory.

An other article : Climatic Sensibility and CO2