Energy-imbalance trends since 2006 fit climate sensitivity estimates consistent with past assessments

Why it matters

The fast rise in Earth’s energy imbalance has raised concern that climate sensitivity is higher than assessed. This analysis finds the trends do not require that. It also shows that the statistical method used to read the data changes the answer a lot, which matters for anyone combining lines of evidence.

What they did

The author used satellite energy-imbalance data (CERES), a forcing ensemble, and the HadCRUT5.1 temperature record to compute trends over 2006–2024. He added model-based estimates of how the warming pattern shifts feedback, and used a Bayesian framework from an earlier assessment. He compared the old likelihood method, which works in temperature, with a revised one that works in energy-imbalance terms, as the energy-balance model itself does. He tested three priors and a case with extra aerosol-cloud forcing.

Key findings

• Taken at face value, the 2006–2024 trends best match a sensitivity of 5.2 K. The older likelihood method peaks at 3.0 K instead.
• With the revised likelihood and a uniform-feedback prior, the posterior median is 3.6 K. Medians run from 3.3 to 4.8 K across the priors.
• Sensitivity below 2 K is strongly disfavored, but the trends say little about the upper bound, which depends on the prior.
• Starting in 2001 gives a higher plug-in estimate of 7.7 K. The author considers this period less reliable because forcing fell before it rose.
• Adding extra aerosol-cloud forcing lowers the uniform-prior median to 2.7 K.

Limitations

• Results rest on one short period, and internal variability could not be ruled out as a cause of the high central estimate.
• Forcing estimates, especially for aerosol-cloud interactions, may be wrong. The pattern-effect adjustment was assumed to be zero, with a large uncertainty.
• This is a preprint that has not been peer reviewed. The process-based prior is illustrative, and the trend evidence was not formally combined with historical evidence.

Glossary

• Equilibrium climate sensitivity: The long-term global warming after atmospheric CO2 doubles and the climate settles.
• Earth’s energy imbalance: The gap between energy Earth absorbs and energy it sends back to space.
• Likelihood: A measure of how well each possible sensitivity value matches the observed data.
• Pattern effect: The change in radiative feedback caused by how warming is spread across the globe.