A Finer Mesh to Improve Cloud Representation in Climate Models?

Bhattacharya, A., Pacific Northwest National Laboratory

Cloud Processes

Cloud Life Cycle

Boutle IA, SJ Abel, PG Hill, and CJ Morcrette. 2014. "Spatial variability of liquid cloud and rain: observations and microphysical effects." Quarterly Journal of the Royal Meteorological Society, 140(679), 10.1002/qj.2140.


Different sizes of water droplets as well as varying water content dramatically alter cloud properties—often at a resolution finer than is currently in use by most climate models.


Different sizes of water droplets as well as varying water content dramatically alter cloud properties—often at a resolution finer than is currently in use by most climate models.

Although clouds can extend for several kilometers, their properties—for example, liquid and rainwater content—can change dramatically over very short distances: far shorter than the 40-by-40 kilometer 'grid' resolution that climate models currently use.

Scientists find it challenging to accurately represent clouds in climate models. As a result, predictions of cloud occurrence, and eventually rainfall, can be often inaccurate or have large biases, i.e., models can over- or under-estimate rainfall amounts by up to a factor of five.

It is no easy task matching the short spatial scales of changing cloud properties with those of other weather parameters, like wind speeds, that vary over larger distances.

Researchers from the United Kingdom Meteorological Office have suggested a way forward, something they argue may be a first, in a paper they published last month in the Quarterly Journal of the Royal Meteorological Society.

Combining observations from three different sources—on location from research aircrafts, land-based sensors, and space-borne remote sensors—the research team found that the liquid and rainwater content of clouds varies depending on cloud cover itself. Variability is less in overcast regions but more in broken cloud fields.

To improve how climate models incorporate such dynamic cloud behavior, the team developed a new parametrization scheme. In other words, based on these observation sets, they developed a method to account for the changing water and rain content of cloud parcels at sub-grid resolution.

This approach resulted in a better estimate of the impact of cloud properties on climate models, and the authors feel confident that such detailed knowledge of cloud and rain content will in fact improve representation of the cloud systems in climate models.