Evolution of Organized Shallow Convection

Uncertainty in the response of clouds to warming is the leading source of uncertainty in projections of future warming. To a large fraction the frequently occurring shallow cumulus clouds in the trade wind region contribute to this uncertainty. In symbiosis with thin clouds of stratiform extent they often create various cloud patterns.

We introduce a neural network that is able to detect the mesoscale organization from GOES16 and MODIS satellite imagery in order to put eight years of ground-based measurements of the Barbados Cloud Observatory into the context of mesoscale organization. With this combination of long-term ground-based measurements from the trade-wind region and satellite image classifications, we overcome the common resolution limitations of satellite derived cloud products of shallow cumuli and are able to present the characteristics of shallow convection depending on the mesoscale organization with great detail.

By using back-trajectories and EUREC4A field campaign data, we show that differences in the atmospheric environment are not only present at the time of pronounced mesoscale organization, but are already distinguishable days ahead in LTS, wind speed and SST.