Sensitivity of Cloud Optical Depth Retrieved from Zenith Radiance to Cloud Phase Assumptions

Cloud optical depth is an important cloud property in climate simulation, yet tends to be poorly predicted. If we are to make progress, we need high-quality retrievals of cloud optical depth for verification purposes. Such measurements are routinely made by satellites, but there is a notable lack of locations where routine ground-based observations of cloud optical depth are made.

The worldwide sunphotometer network of NASA AERONET allowed an increase in global coverage. Although primarily designed to measure aerosol optical properties, these radiometers have been routinely observing cloud properties since 2010, whenever the Sun is obscured by cloud (i.e., cloud mode operation). Retrievals of cloud optical depth are made using zenith radiances at 440 nm and 870 nm, capitalising in strong surface albedo contrasts from vegetated surfaces. However, as the radiance measurements do not contain information on the height of the cloud, we must make the assumption that all cloud optical depth is due to the presence of liquid cloud.

In this presentation, we investigate the potential errors that could arise from this assumption when the cloud profile contains mixed-phase cloud or ice cloud. We focus on observations at the ARM Southern Great Plains site, which has the most extensive ground-based suite of remote sensing measurements to help identify the presence of ice clouds. Using radar retrievals made by the European Project Cloudnet and ACTRIS (Aerosols, Clouds, and Trace gases Research Infrastructure Network), we estimate the frequency of occurrence of mixed-and ice-phase clouds, and quantify their relative contributions to total optical depth. This information is then applied to AERONET Cloud Mode data from the period May 2007 to June 2008 to demonstrate how sensitive cloud mode retrievals are to the cloud phase assumption.