Aerosols, clouds
& climate.

The seed inside the droplet

Water vapour does not simply become a cloud wherever there is dust. A cloud forms when air motion, cooling, moisture and suitable particles meet.

As moist air rises and cools, it can become supersaturated: its water-vapour content exceeds equilibrium with a flat surface of liquid water at that temperature. Some aerosol particles then activate into growing cloud droplets. We call particles capable of doing so cloud condensation nuclei, or CCN, at a stated supersaturation.

Particle size and soluble material influence activation. Dissolved substances lower water activity, favouring water uptake; the curvature of a small droplet raises its equilibrium vapour pressure, opposing growth. Köhler theory brings these effects together. A given particle activates only when the environmental supersaturation exceeds its critical value. Explore CCN physics ↗

Why more particles does not mean a rainfall forecast

Warm rain develops when some droplets become large enough to collect others efficiently. A change in the initial droplet population can alter collision and coalescence. But a real cloud also entrains surrounding air, evaporates, rises, mixes and sometimes produces ice.

Those responses can change cloud water, lifetime and extent. Updraught speed affects the supersaturation reached and therefore how many particles activate. Clouds can also modify the aerosol population through chemistry and removal. Cause and response run in both directions.

That is why “more aerosol means less rain” is an unreliable rule for a particular storm or region. A defensible attribution must distinguish aerosol changes from meteorology and follow the relevant cloud regime. A droplet-size experiment is a way to understand one mechanism, not a monsoon prediction.

High clouds introduce ice

Cirrus clouds contain ice crystals, often near the cold upper troposphere. Their formation involves ice supersaturation and temperature-dependent nucleation pathways. Some aerosol particles can promote ice formation heterogeneously; sufficiently cold solution droplets can freeze homogeneously.

Ice crystals are not spherical water droplets. Their shapes influence scattering and polarisation, making depolarisation measurements useful to lidar studies. High, thin clouds can be difficult to see from the ground while still affecting radiation.

Cirrus both reflect solar radiation and interact with outgoing terrestrial infrared radiation. The balance depends on height, optical thickness, particle properties and the underlying scene. The familiar liquid-cloud picture therefore cannot simply be transferred unchanged to all high clouds.

Follow cirrus through the night

An individual observing night supplies the physical picture behind a long-term cloud statistic. At Gadanki, the lidar record below follows a high cloud system from evening to the following morning.

Read each vertical slice as the atmosphere observed above the station at that time. The changing cloud boundary includes air moving through the beam; it is not a track following one ice crystal.

Sixteen years looking above Gadanki

At Gadanki in southern India, Pandit and colleagues used a ground-based lidar record from 1998–2013 to study cirrus height, geometrical thickness and optical properties. They compared it with CALIOP observations from June 2006 to December 2013: one instrument looking upward from the ground, another downward from orbit. Read the study ↗

Such a comparison asks more than whether two curves resemble each other. Low cloud can obscure a ground lidar’s view of high cirrus. Rain and observing schedules affect which nights enter the record. An orbiting lidar samples at different times and along narrow tracks.

Cloud occurrence in the available profiles is therefore an occurrence relative to the observations made. It should not automatically be called the all-weather fraction of the sky covered by cirrus. The sampling denominator is part of the scientific result.

The panels compare distributions rather than a continuous track through one cloud. Base, top and mid-cloud altitude describe different boundaries; geometrical thickness is the distance from base to top. Relative frequency is normalised to the corresponding sampled population.

A climate question with several moving parts

The IPCC’s Sixth Assessment attributes a net cooling influence to human-caused aerosol changes over the industrial era, including interactions with radiation and clouds. Uncertainty in the size of that influence reflects both aerosol properties and cloud responses. This global assessment does not specify the effect of one particle type everywhere.

In the upper monsoon atmosphere, the questions change again. A 2024 analysis of a Hyderabad balloon observation on 23 August 2017 examined how waves and hydration associated with Typhoon Hato helped explain a tropopause cirrus case. It combined several kinds of evidence rather than assigning cloud formation to aerosols alone. Read the study ↗

To connect a cloud to climate, first identify what changed, what was held fixed and which responses were allowed. That discipline connects a simple droplet experiment with a complex atmospheric model.