Climate scientists have long thought that low clouds, which act like a parasol and cool the earth’s surface, will diminish as the earth heats up – thus amplifying warming in a positive feedback process. This notion has been reinforced by climate models. But recent empirical observations refute the idea and show that the mechanism causing the strongest cloud reductions in models doesn’t actually occur in nature.

The observations were reported in a 2022 paper by an international team of French and German scientists. In a major field campaign, the team collected and analyzed observational data from cumulus clouds near the Atlantic island of Barbados, utilizing two research airplanes and a ship. Barbados is in the tropical trade-wind region where low-level cumulus clouds are common.

More than 800 probes were dropped from one plane that flew in circles about 200 km (120 miles) in diameter at an altitude of 9 km (6 miles); the probes gathered data on atmospheric temperature, moisture, pressure and winds as they fell. The other plane used radar and lidar sensors to measure cloudiness at the base of the cloud layer, at an altitude of 0.8 km (2,600 feet), while the ship conducted surface-based measurements.

The response to global warming of small cumulus clouds in the tropics is critically dependent on how much extra moisture from increased evaporation of seawater accumulates at the base of the clouds.

In climate models, dry air from the upper cloud layer is transported or entrained downward when the clouds grow higher and mixes with the moister air at the cloud base, drying out the lower cloud layer. This causes moisture there to evaporate more rapidly and boosts the probability that the clouds will dissipate. The phenomenon is known to climate scientists as the “mixing-desiccation hypothesis,” the strength of the mixing mechanism increasing with global warming.

But the observations of the research team reveal that the mixing-desiccation mechanism is not actually present in nature. This is because – as the researchers found – mesoscale (up to 200 km) circulation of air vertically upward dominates the smaller-scale entrainment mixing downward. Although mesoscale circulations are ubiquitous in trade-wind regions, their effect on humidity is completely absent from climate models.

The two competing processes are illustrated in the figure below, in which M represents mixing, E is downward entrainment, W is mesoscale vertical air motion, and z is the altitude; the dashed line represents the trade-wind inversion layer.

That cloudiness increases with mixing can be seen from the next figure, which shows the research team’s observed values of the vertical mixing rate M (in mm per second) and the cloud-base cloudiness (as a percentage). The trend is clear: as M gets larger, so does cloudiness.

The research has important implications for cloud feedback. In climate models, the refuted mixing-desiccation mechanism leads to strong positive cloud feedback – feedback that amplifies global warming. The models find that low clouds would thin out, and many would not form at all, in a hotter world.

Analysis of the new observations, however, shows that climate models with large positive feedbacks are implausible and that a weak trade cumulus feedback is much more likely than a strong one. Climate models with large trade cumulus feedbacks exaggerate the dependence of cloudiness on cloud-base moisture compared with mixing, as well as overestimating variability in cloudiness.

Next: Nitrous Oxide No More a Threat for Global Warming than Methane