Tropical clouds, circulation and climate sensitivity

Clouds play an important role in the climate system, as they contribute to the regulation of Earth’s energy budget linked to the amount of solar radiation trapped or reflected back to space (Figure 1.4.10). In general, high, thin clouds at several kilometres altitude have a two-fold warming effect on the climate: They have a high transmissivity for shortwave radiation (incoming sunlight) and low emissivity for longwave radiation (heat), meaning they allow most of the sunlight to reach the surface but block some of the heat escaping to space. In contrast, low, thick clouds reflect more sunlight, and also have a high emissivity for long-wave radiation, allowing more heat to escape, and so have a cooling effect. A changing climate, which causes changes in temperature, humidity and circulation patterns, affects the formation and dynamics of these clouds. This, in turn, can influence the climate and how much warming results from increased atmospheric CO2 concentrations (i.e. ‘climate sensitivity’).

Figure: 1.4.10
Figure 1.4.10: The role of clouds in regulating Earth’s radiation budget

Evidence for tipping dynamics

Literature on cloud-induced tipping points is very limited. Yet cloud-forming processes exhibit strong hysteresis on weather timescales. Indeed, a cloud droplet forms when water starts to stick to a particle after a certain level of humidity (in which a so-called hygroscopic aerosol particle crosses a humidity tipping point into an unstable condensational growth phase); and precipitation, once initiated, is a self-reinforcing cascade where larger particles fall faster and hence grow faster by collisions. Coupling of these micro-scale processes to atmospheric dynamics can lead to spontaneous and irreversible transitions at the intermediate mesoscale – in particular, the transition of shallow cloud layers from closed to open-cell geometries (honeycomb-like cloud patterns formed by convecting air) (Feingold et al,. 2015) and self-aggregation of deep convection (Muller et al,. 2022). Both of these significantly decrease cloud cover and albedo, potentially enabling climate interactions. Could further coupling out to planetary scales produce climate-relevant tipping behaviour? Complicating this question is the fact that cloud-related processes are not well represented in current climate models, limiting their ability to guide us.

The most-discussed possibility has been the extreme case of a global climate runaway. If the atmosphere became sufficiently opaque to infrared (i.e. if it became harder for longwave heat energy to escape due to overcast high cloud, very high humidity, or CFC-like greenhouse gases filling in spectral absorption windows), the planet could effectively lose its ability to cool to space, producing a Venus-like runaway. Although general circulation models (GCMs) and palaeoclimate evidence suggest climate sensitivity rises as climate warms (Sherwood et al., 2020), calculations show virtually no chance of runaway warming on Earth at current insolation levels (Leconte et al,. 2013).

A more plausible scenario is unexpectedly strong global positive amplifying radiative feedback from clouds and high climate sensitivity. Although presumably reversible, this would be serious. With respect to high clouds, suggested missing feedbacks (due to novel microphysical or aggregation mechanisms) have generally been negative (e.g. Mauritsen and Stevens 2015). Low clouds are a greater concern: one recent study using a multiscale atmospheric model found a strong and growing positive amplifying feedback from rapid disappearance of these clouds (Schneider et al,. 2019), highlighting the possibility of nonlinear cloud behaviour and surprises (Bloch-Johnson et al.. 2015, Caballero and Huber, 2013). Although various observations generally weigh against high-end climate sensitivities above 4oC per CO2 doubling, they cannot rule them out (Sherwood et al., 2020).

A final possibility is surprising reorganisations of tropospheric circulation (i.e. in the lowest layer of the atmosphere). Innovative atmospheric models (Caballero and Carlson 2016; Seeley and Wordsworth 2021) and geologic evidence (Tziperman and Farrell, 2009; Caballero and Huber 2010) have suggested possible ‘super-MJO’ (the ‘Madden-Julian Oscillation’ being the dominant mode of ‘intraseasonal’ variability in the tropical Indo-Pacific, characterised by the eastward spread of enhanced or suppressed tropical rainfall lasting less than a season) and/or reorganisation of the tropical atmospheric circulation in a warmer climate due to cloud-circulation coupling. These scenarios are supported by little evidence, but if they did occur they could massively alter hydrology in many regions. Poor representation of tropical low clouds has also likely inhibited coupled model simulations of decadal variability or regional trends (Bellomo et al,. 2014; Myers et al. 2018), raising the possibility that, even if clouds cannot drive tipping points, they might amplify other tipping points in ways that are missing from current models.

Assessment and knowledge gaps

In summary, concern about cloud-driven tipping points is relatively low. Cloud feedbacks will, however, likely affect the strength of climate responses, including for many tipping points. For example, they could potentially amplify variability, and current models may not be capturing this well. High climate sensitivity from strongly positive cloud feedbacks also cannot be ruled out.

Bezos Earth Fund University of Exeter logo
Earth Commission Systems Change Lab logo Systemiq logo
Global Tipping Points logo
Share this content