| Summary: | Dissertation (Ph.D.) University of Alaska Fairbanks, 2000 Use of satellite data to study the surface and cloud properties and the solar radiation budget (SRB) is very important for improving our understanding of cloud and sea-ice albedo feedback in the Arctic. Based on an accurate and comprehensive Radiative Transfer Model (RTM), algorithms were developed for using the National Oceanic and Atmospheric Administration (NOAA) Advanced Very High Resolution Radiometer (AVHRR) data for the discrimination of cloud from snow/ice surfaces, retrieval of snow surface properties and surface albedo, and retrieval of cloud optical depth (tau) and effective droplet size ( re). Through the improved estimation of solar reflectance in AVHRR channel 3 (3.75 mum) and atmospheric anisotropic correction, a threshold function was found and used for developing an automatic cloud discrimination algorithm over snow/ice surfaces. Thin cirrus was discriminated using the brightness temperature difference between AVHRR channels 4 and 5 and brightness temperature in channel 4. Retrieval of snow grain size and mass-fraction of soot from AVHRR is difficult because of the effects of aerosol in channel 1 and the strong water vapor absorption in channel 2. Retrieval of surface albedo is more promising, but, with the melt of snow/ice, different narrow-to-broadband conversion relations should be used to derive broadband albedo. AVHRR channels 2, 3 and 4 are used to retrieve tau, r e and cloud top temperature simultaneously. Validation of these algorithms with in-situ aircraft measurements by the NCAR C-130 and the NASA ER-2 and with surface measurements obtained during the Surface Heat Budget of the Arctic Ocean (SHEBA) experiment indicates that the retrieved re is close to the "true" value of re, but the retrieved tau tends to be overestimated. Uncertainties of cloud retrievals with regard to cloud cover fraction, vertical inhomogeneity, multi-layer stratification and cloud phase were examined. Inter comparison of different satellite data ...
|
|---|