(Graphique) C. Bickel /Science; (Données) Meehl et al., Sci. Adv., 6, eaba1981 (2020)
Deuxièmement, l'équipe a sondé les rétroactions climatiques individuelles. Certains de ces effets, comme l'effet de réchauffement de la vapeur d'eau, sont bien connus. Mais les nuages, qui peuvent refroidir ou réchauffer la planète en fonction de la façon dont ils réfléchissent la lumière du soleil et piègent la chaleur, ont longtemps été un joker. En particulier, les climatologues veulent comprendre les ponts des nuages stratocumulus qui se forment au large des côtes. S'ils se développent plus étendus en réponse au réchauffement, comme certains le soupçonnent, ils pourraient avoir un effet de refroidissement.
Several years ago, a suite of high-resolution cloud models identified two feedbacks that would have the opposite effect, clearing clouds and worsening warming. In models, higher temperatures allowed more dry air to enter thin clouds from above, preventing them from thickening. At the same time, more CO2 The levels trap heat near the cloud tops, easing the turbulence that causes more clouds to form. Since then, satellites have observed this dynamic in warmer-than-average parts of the atmosphere. “There is a growing consensus that the [cloud] the feedback is positive, but not very important, ”says Thorsten Mauritsen, climatologist at Stockholm University.
Finally, the team looked at records from two past climates – 20,000 years ago, at the height of the last Ice Age, and a warm period 3 million years ago, the last time atmospheric CO2 the levels were similar to today. Recent work suggests that climate sensitivity is not a fixed property of the planet, but changes over time. During warmer periods, for example, the absence of ice caps likely increased sensitivity. Records of old temperatures and CO2 The levels allowed the team to identify sensitivities of 2.5 ° C and 3.2 ° C for cold and hot periods, respectively. “It’s really comprehensive,” says Jessica Tierney, a paleoclimatologist at the University of Arizona, who was not part of the report. Even for the coldest climatic state, she says, the possibility of a sensitivity below 2 ° C seems negligible.
Gathering the three pieces of evidence was a huge task. But connecting them together for a unified prediction was even more difficult, Marvel says. The team used Bayesian statistics to walk through their assembled data, which allowed the researchers to test how their assumptions influence the results. “The real benefit” of Bayesian statistics, says Tierney, is how they allow uncertainties at each stage to fuel an end result. The co-writers have often hit each other’s heads, Marvel says. “It was such a long and painful process.” The final range represents a confidence interval of 66%, corresponding to the traditional “probable” range of the IPCC. The WCRP team also calculated a 90% confidence interval, which ranges from 2.3 ° C to 4.7 ° C, leaving a slight chance of warming above 5 ° C.
Either way, the report has a simple conclusion, Sherwood says: a doubling of CO2 anything but guarantees warming of more than 2 ° C. “Three main lines of evidence are all very difficult to reconcile with the lower end of climate sensitivity.”
In recent years, another uncertainty in the future of the climate has also narrowed: Global emissions appear unlikely to reach the worst-case scenarios the IPCC helped craft 15 years ago, ruling out some forecasts of extreme warming. . “We are light years ahead of where we were in 1979,” says Reto Knutti, co-author and climatologist at ETH Zurich.
Unfortunately, the years of work required to achieve this certainty have come at a cost: 4 decades of additional emissions and relentless global warming.