Search

Monitoring and estimating potential snow disasters in pastoral areas of northern Xinjiang Province are important for decision-making in hazard reduction and prevention. In this paper, four scenes of NOAA/AVHHR (Advanced Very high Resolution Radiometer) images were combined with ground observation data in the north of Xinjiang Province to establish a model for monitoring snow depth. Using a linear spectral decomposition method, the pixel-based snow coverage and snow classification were studied. The spatial characteristics of snow, grassland, animal and climate factors were used to develop two new quantified indices for estimating the potential snow hazard grade and for integrated evaluation of snow disasters to grassland animal husbandry. The criteria for snow hazard grade and snow disaster evaluation were established. Results indicated that: (i) a pixel-based index K1, based on grassland yield, animal capacity, utilisable grassland area coefficient and seasonal grazing utilisation scenarios, can be quantitatively integrated to reflect the grassland capability of resisting snow disasters; (ii) the snow hazard index (K) systematically expresses the spatial and temporal changes of grassland and snow cover, and analyses, and predicts and evaluates the snow hazard grade under conditions where climatic and animal husbandry information may be unavailable during snow disasters. This index plays an important role in studies on early warning of snow hazard in pastoral areas; (iii) the integrated snow disaster evaluation index (E) and related classification criteria reflect the details of snow disaster magnitude in temporal and spatial scales, which provide the basic information for dynamic monitoring and integrated evaluation on snow disasters in pastoral areas. [ABSTRACT FROM AUTHOR]