Moderate Sub-pixel Snow Mapping Algorithm On Tibetan Plateau

Snow is an important component of the Earth's surface. Because of its importance from both a scientific and resource management standpoint, accurate monitoring of snow cover extent is an important research goal in the science of Earth systems. The Tibetan Plateau is the most sensitive area in the world to climatic change. Moderate resolution image is currently the most suitable data for monitoring regional snow extent. The high spatial resolution and numerous MODIS spectral bands in the 0.4 to 2.5 μm wavelength region allow more accurate monitoring of snow cover extent on a global basis than is possible with other operational satellites.The goal of this dissertation is to investigate novel methods of remote sensing technologies to improve the accuracy of mapping snow cover. Medium spatial resolution remotely sensed imagery is comparatively very cheap, but has a critical drawback "mixed" pixels. The problem is severe over rugged terrain, where extreme variations in snow cover, vegetation type, canopy density, and lithology occur over small horizontal distances, leading to mis-classification of snow covered areas.Mixture modelling is becoming an increasingly important tool in the remote sensing community as researchers attempt to resolve the sub-pixel, mixture information, which arises from the overlapping land cover types within the pixel's instantaneous field of view. This paper described and summarized all kinds of mixture models. Selecting spectral endmembers is the key in a spectral unmixng algorithm. In current studies about snow spectral unmixng algorithm, the major difference is the techniques in selecting spectral endmembers. The analyses of the current techniques in selection of the reference spectral endmembers is also done in this papar.In the paper, I tested four different sub-pixel analysis methods: Linear Mixture Model (LMM), Fuzzy c-means Clustering (FCM), Back-Progagation Neural...