Abstract
Ground-based observations and satellite data have been compared for the 18 October 1989 case study of the International Cirrus Experiment (ICE) field campaign. They correspond to thin cirrus clouds with infrared emittances in the range 0–0.3. Good correspondence was obtained when comparing the time variability of the effective downward beam emittance of the cirrus clouds observed at Nordholz (53.8°N, 8.3°E) to the spatial variability of the effective upward beam emittance derived from NOAA-11 Advanced Very High Resolution Radiometer (AVHRR) data acquired at 1225 UTC. A simple model of cirrus cloud particles was found to satisfy both the ground-based observations of the angular dependence of the scattered solar radiation at 0.85 µm and the satellite observations of the brightness temperatures in channel 4 (11 µm) and channel 5 (12 µm) of NOAA-11 AVHRR. The best fit was obtained for fully randomly oriented hexagonal ice plates with a thickness of 10–20 µm and a diameter of 200–500 µm. Although actual cloud ice crystals are probably not all hexagonal plates, our simple model of randomly oriented ice plates allows us to appropriately simulate the optical properties of the observed cirrus in which particles surely present a large variety of shapes. The equivalent radius of the retrieved ice plates (i.e., the radius of spheres of the same volume) is 50–80 µm. However, ice spheres do not simulate the halo of cirrus clouds observed from the aureolemeter measurements. Moreover, assuming spherical particles to explain brightness temperature measurements in AVHRR channels 4 and 5 leads to an effective radius of 27 µm, which is noticeably smaller than the one obtained with the hypothesis of hexagonal plates.
On the other hand, analysis of AVHRR data also highlights the important difference between natural thin cirrus and jet contrail microphysics. Contrails are revealed to be composed of smaller equivalent spherical particles with an effective radius of about 4.5 µm.