This paper investigates the ability of the Information Theoretic Snow Detection Algorithm (ITSDA) in detecting changes due to snow cover between summer and winter seasons on large area images acquired by COSMO-SkyMed constellation. ITSDA is a method for change detection in multitemporal SAR images, which has been recently applied by the authors to a subset of Cosmo-SkyMed data. The proposed technique is based on a nonparametric approach in the framework of Shannon’s information theory, and in particular it features the conditional probability of the local means between the two images taken at different times. Such an unsupervised approach does not require any preliminary despeckling procedure to be performed before the calculation of the change map. In the case of a low quantity of anomalous changes in relatively small-size images, a mean shift procedure can be utilized for refining the map. However, in the present investigation, the changes to be identified are pervasive in large size images. Consequently, for computational issues, the mean shift refinement has been omitted in the present work. However, a simplified implementation of mean shift procedure to save time will be possibly considered in future submissions. In any case, the present version of ITSDA method preserve its characteristics of flexibility and sensibility to backscattering changes, thanks to the possibility of setting up the number of quantization levels in the estimation of the conditional probability between the amplitude values at the two acquisition dates.