| Popis: |
The trades and the inherent trade cumulus clouds cover large parts of the tropicaloceans. Trade cumulus clouds are ubiquitous but also very small in their horizontaland vertical extent posing huge challenges on observing systems such as satelliteimagers. Climate models exhibit a significant spread in the response of tradecumulus clouds to global warming motivating their intense study in recent years.Within this thesis, I use high-resolution satellite images to gain new insights onsmall and optically thin clouds in the trades.The way trade wind clouds change with surface warming is decisive for theirfeedback, which defines whether clouds further amplify or dampen the warmingof the climate system. Cloud feedback estimates can be investigated from so-calledcloud-controlling factors, their relation to cloud properties in the current climateand their change with global warming. Results from my first study indicate awind-speed driven boundary layer in the trades. The surface trade winds showthe most powerful control on cloud properties such as cloud sizes, top heights orcloud clustering. Furthermore, the Bowen ratio was firstly tested from observationsand emerges as a potential new control factor. Trade cumulus cloud properties alsoshow a susceptibility to the sea surface temperature and the stability of the lowertroposphere which are both projected to change in a warming climate and maythus impact cloud feedbacks.Investigating cloud-controlling factors is an ongoing task and seems to be withinreach from extensive measurements of the recent field campaign EUREC4A. Firstanalysis of cloud observations from multiple instruments indicate the frequentoccurrence of not only small, but also optically thin clouds. Due to their lowreflectance, such clouds are challenging to detect from passive imagers. High-resolution imagers are able to detect small clouds, but, do conventional satellitecloud products still miss optically thin clouds?Within another study, I follow a new approach for defining the total cloudcover consisting of clouds detected by conventional cloud masking schemes andof undetected optically thin clouds. By simulating the well-understood clear-skysignal I can extract clouds as a residual from the all-sky observation and circumventconventional but problematic thresholding tests in cloud masking schemes. Fromevaluating a high-resolution satellite dataset collected during EUREC4A, I findthat optically thin clouds contribute 45 % to the total cloud cover and reducesthe average cloud reflectance by 29 %. Undetected optically thin clouds can havemajor implications for estimates of the radiative effect of clouds and thus, cloudfeedbacks. |
