Close this search box.

New Data Imply Slower Global Warming

London, 24 April — A paper just published by the Journal of Climate concludes that high estimates of future global warming from most computer climate simulations are inconsistent with observed warming since 1850. The implication is that future warming will be 30 to 45% lower than suggested by the simulations.


The study estimates climate sensitivity — how much the world will warm when carbon dioxide levels increase* — from changes in observed temperatures and estimates of the warming effect of greenhouse gases and other drivers of climate change, from the mid/late 19th century until 2016.

The paper also addresses previous criticisms of the methodology used, finding that these are unfounded.

Nicholas Lewis explains,

“Our results imply that, for any future emissions scenario, future warming is likely to be substantially lower than the central computer model-simulated level projected by the IPCC, and highly unlikely to exceed that level.”

Nicholas Lewis adds,

 “Our new sensitivity estimates are slightly lower than those obtained in a predecessor study published several years ago, despite the inclusion of the strong 2015–16 El Niño warming. Importantly, the upper uncertainty bounds of the new estimates are much lower.”


  • The estimates of effective radiative forcing given in the latest IPCC Assessment Report (AR5) are used, extended up to 2016, with recent revisions to greenhouse gas forcing and post-1990 ozone and aerosol forcing estimates incorporated.
  • A median estimate for ECS* of 1.66°C (5–95% uncertainty range: 1.15–2.7°C) is derived using globally-complete temperature data. The comparable estimate for 31 current generation (CMIP5) computer climate simulation models is 3.1°C.
  • One of the chief criticisms of the method used is that it does not allow for the possibility of climate sensitivity varying with time after imposition of forcing, as it does in most CMIP5 models. However, the paper shows that when calculated so as to correctly reflect CMIP5 models’ behaviour, this possible effect is immaterial to either the study’s or CMIP5 models’ median ECS estimates.§
  • A median estimate for TCR* of 1.33°C (5–95%:1.15–1.9°C) is derived using the same data. The comparable estimate for the 31 CMPI5 models is 1.9°C.
  • The estimates of climate sensitivity are remarkably insensitive to the period of analysis chosen (see Figure 1).

Figure 1: Remarkably, estimation of climate sensitivity using the 1927–41 mean gives the same sensitivity estimate as using the 2007–16 mean. Circles show 1872-2016 pentadal-mean changes in net outgoing radiation (cyan pre-1927, blue post-1927) plotted against change in globally-complete surface temperature. The slope of the relationship is inversely proportion to climate sensitivity. Red squares show 15-year means from 1927-41 to 2007-16. Despite the effects of multidecadal internal variability, all the 15-year means lie close to the 1872-2016 pentadal-means regression best-fit line.

  • These median ECS and TCR estimates imply multicentennial or multidecadal future warming under increasing forcing of only 55−70% of the central warming projection using CMIP5 models.
  • It has been suggested in various studies that forcing-efficacy effects (principally, cooling aerosol forcing having a stronger than normal effect), variability in sea-surface warming patterns and temperature estimation issues likely lead to climate sensitivity estimates based on warming over the last circa 150 years being biased low. All these issues are examined in detail in the paper, the conclusion being that very minor or no bias was to be expected when using globally-complete temperature data.


Nicholas Lewis and Judith Curry, 2018: The impact of recent forcing and ocean heat uptake data on estimates of climate sensitivity. Journal of Climate, Early Online Release []



Energy budget estimates of equilibrium climate sensitivity (ECS) and transient climate response (TCR) are derived based on the best estimates and uncertainty ranges for forcing provided in the IPCC Fifth Assessment Scientific Report (AR5). Recent revisions to greenhouse gas forcing and post-1990 ozone and aerosol forcing estimates are incorporated and the forcing data extended from 2011 to 2016. Reflecting recent evidence against strong aerosol forcing, its AR5 uncertainty lower bound is increased slightly. Using a 1869–1882 base period and a 2007−2016 final period, which are well-matched for volcanic activity and influence from internal variability, medians are derived for ECS of 1.50 K (5−95%: 1.05−2.45 K) and for TCR of 1.20 K (5−95%: 0.9−1.7 K). These estimates both have much lower upper bounds than those from a predecessor study using AR5 data ending in 2011. Using infilled, globally-complete temperature data gives slightly higher estimates; a median of 1.66 K for ECS (5−95%: 1.15−2.7 K) and 1.33 K for TCR (5−95%:1.0−1.90 K). These ECS estimates reflect climate feedbacks over the historical period, assumed time-invariant. Allowing for possible time-varying climate feedbacks increases the median ECS estimate to 1.76 K (5−95%: 1.2−3.1 K), using infilled temperature data. Possible biases from non-unit forcing efficacy, temperature estimation issues and variability in sea-surface temperature change patterns are examined and found to be minor when using globally-complete temperature data. These results imply that high ECS and TCR values derived from a majority of CMIP5 climate models are inconsistent with observed warming during the historical period.


*Two standard metrics summarize the sensitivity of global surface temperature to an externally imposed radiative forcing. Equilibrium climate sensitivity (ECS) represents the change in temperature to a doubling of atmospheric CO2 concentration once the deep ocean has reached equilibrium. The transient climate response (TCR), a shorter-term measure over 70 years, represents warming at the time CO2 concentration has doubled when it is increased by 1% a year.

Changes are relative to the 1850–1884 mean (this period has the same mean volcanic forcing as for 1850-2016). All volcanic forcing has been scaled by 0.55 to adjust for its low efficacy [discussed in the paper]. ΔR (= ΔF − ΔN) is estimated by scaling ΔF, based on ΔF and ΔN values for 2007-16.

PDF copies of a version of the accepted manuscript for the paper and its Supporting Information, along with an article giving information about the paper and its findings, is available on Nicholas Lewis’s personal web-site, here. A blog-post version of the article is available at Judith Curry’s web-blog, here. Nicholas Lewis is the sole or lead auther of six previous peer-reviewed papers about climate sensitivity


Nicholas Lewis

Bath, UK. [email protected]

Judith Curry

Climate Forecast Applications Network, Reno, NV, USA.  [email protected]