Cloud Detrainment

Detrainment is convection’s output to its surroundings: air, moisture, and condensate leaving an updraft into the environment. It is the counterpart of entrainment, which is environmental air drawn into the plume, and the two together determine what a convective cloud does to the atmosphere around it rather than merely within it.

Most detrainment happens near the plume’s level of neutral buoyancy, where the updraft decelerates and spreads laterally. That is why anvils form where they do. But detrainment also occurs throughout the depth of the cloud, and this distributed component is the one that matters most for climate, because it is how convection moistens the mid-troposphere.

Why it carries so much weight

The mid-troposphere is dry, radiatively important, and difficult to moisten by any other route. Large-scale ascent is too slow, and boundary-layer mixing does not reach. Convective detrainment is the dominant supply, which places it upstream of several things that matter a great deal:

  • Mid-level humidity, which controls longwave cooling and therefore the atmosphere’s radiative balance.
  • Subsequent convection. Detrained moisture reduces the entrainment penalty for later plumes rising through the same air. Convection preconditions its own environment, and this feedback is central to how convective organisation develops.
  • Anvil cirrus. Detrained condensate at the equilibrium level forms extensive high cloud whose radiative effect can rival the convection’s own.

The consequence is that a convective scheme’s detrainment assumptions propagate well beyond precipitation. A scheme that detrains too little leaves the mid-troposphere too dry, which suppresses subsequent convection and biases the radiation.

The parameterisation problem

In a cumulus scheme detrainment is not derived — it is specified, usually through an entrainment-detrainment rate profile chosen partly on physical grounds and partly because it produces acceptable results. This is one of the least constrained parts of convective parameterisation, and it is a substantial contributor to the spread in climate sensitivity across models, since it sets both mid-level humidity and high cloud amount.

It also sets convective cloud fraction, since detrained condensate is what the cloud fraction diagnostic sees. Radiation, cloud fraction, and detrainment are a single coupled choice rather than three independent ones, which is why changing a convection scheme without revisiting the radiation configuration tends to produce energy-budget surprises.

Convection-permitting models improve on this by resolving the plumes, so detrainment emerges from the dynamics. They do not escape it entirely — detrainment happens at scales near the grid, and at 1-4 km resolution the plumes are marginally resolved at best.

See also: cumulus convection schemes, convective cloud fraction, and radiation schemes for the coupling this sits inside.