Derivation of Vertical Profiles of Droplet Size in Cumulus Clouds from Passive Remote Sensing Observations by the Research Scanning Polarimeter

The Research Scanning Polarimeter (RSP) is an airborne along-track scanner measuring the polarized and total reflectance in 9 spectral channels. Its uniquely high angular resolution allows for characterization of liquid water cloud droplet sizes using the rainbow structure observed in the polarized reflectance over the scattering angle range from 135 to 165 degrees. Such an angular resolution coupled with high frequency of the RSP measurements also allows for geometric constraint of the cumulus cloud's 2D cross section between a number of tangent lines of view, thus, providing estimates of the macroscopic parameters of the cloud, such as its geometric shape, dimensions, and height above the ground.

The rainbow is dominated by single scattering of light by cloud droplets, so its structure is characteristic specifically of the droplet sizes at cloud surface (within unit optical depth into the cloud, equivalent to approximately 50 m).
A parametric fitting algorithm applied to the polarized reflectance provides retrievals of the droplet effective radius and variance assuming a prescribed size distribution shape (gamma distribution). In addition to this, we use a non-parametric method, the Rainbow Fourier Transform (RFT), which allows us to retrieve the functional shape of the droplet size distribution.

While our previous analyses were focused on horizontal variability of cloud droplet size, in this study for the first time we evaluate the possibility to retrieve vertical profiles of microphysical characteristics along the illuminated side of the cumulus cloud. This task requires a combination of our previously developed micro- and macrophysical retrieval methods. First, we use RSP's lines of view tangent to the cloud surface to determine its shape. Then for each point on the bright side of this surface we aggregate view lines passing through it, thus, creating the polarized rainbow corresponding to this point. The structure of this rainbow is analyzed yielding effective radius and variance of the droplet size distribution at this point. Finally, the retrievals made for all available points at the cloud surface are combined into vertical profiles of the droplet effective radius and variance.

We present the results of testing of the proposed profiling algorithm on simulated data. For these tests we modeled clouds using LES and then performed 3D radiative transfer (RT) computations for the resulting cloud field. The RT dataset was then sampled to emulate "virtual RSP measurements" made at certain altitude above the clouds. These "measurements" were analyzed using the proposed technique and the retrieved profiles were compared to the actual data from the LES model output. Cumulus congestus (towering cumulus, Tcu) clouds were selected for LES modeling in preparation for analysis of real data from the upcoming Cloud, Aerosol and Monsoon Processes Philippines Experiment (CAMP2Ex). Such clouds occupy wide altitude ranges, thus, providing good opportunities for testing of our profiling technique.

In addition to tests on simulated data we also present some cloud profiles retrieved from real RSP measurements made during past field campaigns, and we look forward to CAMP2Ex for comprehensive in situ validation of our techniques.