Evaluation of the diurnal evolution of flower cloud organization in multi-day Lagrangian large-eddy simulations based on EUREC4A against GOES satellite data

The subtropical marine stratocumulus-to-cumulus cloud transition and associated cloud size distributions are studied using Large-Eddy Simulations based on EUREC4A data. The simulations with the DALES code follow a Lagrangian trajectory from initially overcast stratocumulus to the tropical shallow cumulus region at the HALO flight site near Barbados, covering four days and three complete diurnal cycles. Mean state and bulk properties for different domain sizes are evaluated against aircraft data. In addition, time-continuous high-frequency data from Geostationary Operational Environmental Satellite (GOES) are used to investigate the evolution of cloud size distributions, focusing on the diurnal evolution of mesoscale cloud features. TOA brightness temperature data from GOES is used at a spatial resolution of 2x2 km² to characterize cloud populations and cloud morphology. These are compared to TOA brightness temperatures calculated from DALES output using the RRTOV simulator, as applied to subdomain-averages of similar dimensions. We find that the simulation with the largest domain size (100x100 km²) best reproduces the observed boundary layer and cloudy states at the HALO target site. The same applies to the amplitude and evolution of cloud cover as detected by GOES. The upstream nocturnal flower cloud organization is also reproduced, albeit with a slight time delay.