Clouds in the extreme: challenging satellite retrievals of atmospheric properties

In light of new satellite technology aimed at the improved quantification of cloud properties, we revisit some of the heritage cloud and cloud forcing products made available over the past 10 years. ISCCP products and MODIS retrievals, in particular, are compared against ship-based cloud and aerosol measurements collected in a wide range of atmospheric and oceanic conditions. Those regions which pose the greatest challenge to, and thus incur the largest uncertainty in, cloud retrieval are targeted for comparison: high latitudes during rapidly changing ice conditions, and aerosol-perturbed environments such as the Mediterranean and Red Seas, and the Saharan Air Layer outflow.

In each of these test cases, it is a combination of atmospheric and surface effects which complicates cloud retrieval. In the polar regions during the daylight season, sea ice and snow melt is often coupled with an increase in cloud cover and a shift towards multiple cloud layers. A significant percentage of these clouds are mixed phase, often low-level, and difficult to distinguish from the ice surface beneath using visible and thermal imagery. In the Mediterranean and Red Seas, the radiative effect of seasonal intrusions of desert dust from Africa and the Arabian Peninsula is comparable to surface cloud forcing in pristine maritime environments. Additional atmospheric contributions from pollution aerosols, particularly in the eastern Mediterranean, foil aerosol speciation and cloud amount retrievals, and thus affect forcing calculations based on remotely-sensed data. Off of West Africa, the competition between aerosol-related scattering and large variations in the atmospheric water vapor content intermingled with cloud cover from mesoscale convective systems and contamination of the signal by sun glint make this region particularly challenging to satellite retrieval, especially for passive and nadir-viewing sensors.

Time series of cloud amount, incident radiation, atmospheric profiles, and aerosol optical depth measured by ship-based instrumentation in each of these regions will be collocated with ISCCP cloud amount, MODIS cloud amount and optical depth, MODIS aerosol optical depth at 550nm and aerosol type retrievals, TOVS and AIRS water vapor content, and multi-sensor cloud forcing calculations. Attention will be particularly paid to differing spatio-temporal measurement scales and the influence of water vapor and aerosol on the cloud retrieval and forcing calculation. The final satellite-in situ differences in cloud cover and forcing are couched in terms of advancing satellite technology and possible error remediation offered by CALIPSO and CloudSat and synergies between the A-Train Satellites.