Thèse (Ph.D.)--Université de Sherbrooke (Canada), 2007. / Titre de l'écran-titre (visionné le 1 février 2007). In ProQuest dissertations and theses. Publié aussi en version papier.

The dielectric properties and emissivity of several types of sea ice were studied to improve upon the currently available algorithms for intepreting polar radiometric imagery from the SSM/I sensor. Laboratory studies indicated that the relative dielectric constant of new sea ice, as measured with a spectral C-Band radiometer, is approximately 12. This value represents the early stages of ice formation, and decreases rapidly with ice growth to approach the accepted value of 3.2 for first year ice. Atmospheric effects on satellite based passive microwave data were also examined with regard to the 19, 22, 37, and 85 GHz channels of the SSM/I system. The atmospheric vapor and cloud liquid water attenuation was empirically modeled with respect to frequency and atmospheric temperature. The mean atmospheric temperature was empirically linked to surface temperature through an exponential relationship. The surface emissivity was then modelled as a function of surface type via a mixing formalism, where the emissivity of open water depends on surface wind-speed. With the significant environmental variables parameterized, two algorithms for analysis of polar SSM/I data were developed. The first is a simple modification that adds the capability of determining ice temperature to an existing algorithm. The second is a weather correcting algorithm significantly more complex than those currently in use for the lower three frequencies of the Special Sensor Microwave Imager (SSM/I). The uncertainty of the six products due to measurement noise was estimated at 5% for surface parameters and 10% for atmospheric variables. The atmospheric vapor and cloud liquid water estimates are used to convert the raw 85 GHz brightness temperatures to surface brightness temperature estimates, producing a significantly enhanced ice edge.