Datensatz

Zusammenfassung

Clouds play a key role in the Earth’s climate system hence understanding fully their driving processes and distribution patterns is of importance to better characterize and model future states of the climate. A common approach to characterize clouds is through cloud regimes. Each type demonstrates very different radiative properties and interacts in various ways with aerosol particles in the atmosphere. Nevertheless, it has proven challenging to characterize cloud regimes objectively through remote sensing data.Building upon the method we previously developed, we combine synoptic observations and passive satellite remote-sensing retrievals to constitute a database of cloud types and cloud properties to eventually train a cloud classification algorithm. Cloud regime labels are provided through the global marine meteorological observations dataset (UK Met Office, 2006) which is comprised of near-global synoptic observations. The cloud classification model is built on different cloud-top and cloud optical properties (Level 2 products MOD06/MYD06 from the MODIS sensor) extracted in the vicinity of the observation. To make full use of the large quantity of remote sensing data available and to investigate the variety in cloud settings, a convolutional variational auto-encoder (VAE) is applied as a dimensionality reduction tool in a first step. The cloud classification task is subsequently performed drawing on the constructed latent representation of the VAE. Using simulation data produced by the ICON global climate model we can further investigate characteristics of the cloud regimes, their representation and distribution.