The satellite observation of drizzle in Stratocumulus clouds

Stratocumulus (Sc) clouds stretch hundreds of kilometers horizontally in the Earth's atmosphere. They usually form at the bottom layer of the troposphere and consist mostly out of spherical, water droplets. They feature approximately horizontally homogeneous structure, in terms of droplet size, concentration and liquid water content. When a large number of such droplets cluster together, they scatter more of the incoming shortwave radiation, rather than absorb it. However, this established radiative character depends on the persistence of the droplet size distribution within the cloud. Once they have grown large enough, via diffusion and coalescence processes (near the droplet radius of 14-16µm), drizzle occurs. It is known to deplete the cloud's liquid water content over a period of several hours, leaving behind a cellular cloud structure (as seen from satellite imagers). This transformation has been modeled and referred to in the literature as the Pockets of Open Cells (POCs).

When does such a cloud completely get depleted? Once the corresponding post-drizzling volume domain of the atmosphere contains nothing but the water vapor in an unsaturated environment, the cloud section will cease to exist. Even though this process lasts over a period of hours, it is challenging to quantify it. As the drizzle is formed at the base, shielded by hundreds of meters thick cloud deck, it is hard to observe it using satellite passive remote sensing and evaluate its impact on the overall cloud lifetime, as well to quantify the local change in the cloud radiative character.

The observations of an imaging radiometer on board the satellite (for example, geostationary satellite such as Meteosat Second Generation) are representative of a far-reaching two-dimensional domain in the horizontal direction, but without the capacity to profile the cloud vertically, throughout the span of the covered domain. Here we use the simulated irradiance calculated at the top of the cloud (as it would be observed by a space-borne passive radiometer) in the near and thermal infra red part of the spectra, as a proxy to determine the stage of droplet growth and an ongoing precipitation process intensity at the cloud base.

We present here an observational technique using simulated satellite imaging radiometry via the framework of the EarthCARE SIMulator (ECSIM). Multiple Spectral Imager instrument model is applied to the ingested cloud scenes of a Dutch Atmospheric Large Eddy Simulator (DALES). A period of 40 hours of the cloud field evolution reconstructed by DALES during the Atlantic Stratocumulus Transition Experiment campaign is used to create a series of cloud scenes of a transitioning Sc into a Sc topped Cumulus (Cu) fields. Drizzle has appeared throughout the cloud evolution, often evaporating on its way to the surface and clearing the cloud.

The brightness temperatures are calculated using a three-dimensional, Monte Carlo, long-wave Radiative Transfer Model (RTM), integrated in the framework of the ECSIM. This simulated Brightness Temperatures Difference (BTD) between the channels at 3.9 and 11μm from space-borne imagers is then used to highlight the cloud droplet size spatial variability, during the evolution of the drizzle near the cloud base. The observed correlation of the BTD with the droplet size variability is used to interpret the conditions at which the precipitation at the cloud base is triggered. Tracking the process of evolution of the cloud droplet into a precipitating drop is possible due to the strong sensitivity of the 3.9μm channel to the particle size and cloud phase. A comparison of the observed BTD with the vertically averaged cloud droplet size from the imported cloud scene is done. It is used to validate if a single pixel value from the cloud top (as retrieved via RTM) can be representative of the inhomogeneous microphysical vertical structure of a Sc deck and the potential precipitation at the cloud base. A range in BTD for the two infra-red channels, between 0 and 2K is correlated to the effective radius of the cloud droplets larger than 15µm. According to the modeling studies this value is found as the threshold of the droplet radius during the growth process that triggers the drizzle initiation.

The drizzle contaminated cloud profiles from DALES have shown a discrepancy to the adiabatic cloud vertical structure, where the largest droplets are concentrated at the cloud top. Drizzling cloud features vertical profiles where the largest droplets are near the cloud base, whilst maintaining the cluster of large droplets at the top, relative to the surrounding horizontal level. This enabled us the usage of a dual-channel BTD retrieval of the cloud droplet size based on the cloud-top irradiance, to determine the precipitation character at the cloud base, for a Sc cloud.