Monitoring Snow Cover In Pastoral Areas On Qinghai-Tibetan Plateau

The Qinghai-Tibetan Plateau(QTP), one of the three snow cover centers as well as the major pastoral area in China, which is situated at 4000 m above sea level averagely, has a complex topography, cold and wet climatic condition. Because of adverse natural conditions and backward economy and management situations, the sustainable development of animal husbandry can be severely influenced by snow-caused disasters, which are major natural hazards in winter and spring seasons in this region. The main goal of this thesis is to understand the spatial and temporal variation of the snow cover on QTP, to establish QTP snow depth model and snow cover monitoring information system, and to provide disaster information and aid decision-making strategies to reduce economic losses of disaster areas by using Remote Sensing (RS), Geographic Information Systems (GIS) and computer network technology and analysing the MODIS and AMSR-E data.Daily in situ measurements of snow depth data at 106 meteorological stations and the MOD10A1 and MYD10A1 data in six snow seasons from November 1 to March 31 of 2002 to 2008 were collected and used to evaluate the snow classification accuracy of NASA EOS MODIS snow products on QTP; New daily,5-day,10-day snow cover products (500 m) were developed through combining MODIS daily snow cover data and AMSR-E daily snow water equivalent(SWE) data to analyse the snow cover spatial and temporal variation in the study area; Landsat-ETM images and MODIS surface reflectance product (MOD09GA) were used to find out the best Normalized Difference Snow Index(NDSI) threshold of QTP, and then the new snow mapping method were used to evaluate the snow classification accuracy. Daily snow depth, precipitation, maximal temperature, minimum temperature, average temperature, wind velocity and snow duration data recorded by meteorological stations and AMSR-E passive microwave Brightness Temperature(BT) data during the 7 snow seasons from 2002 to 2009 on QTB were used to establish the snow depth model, and then the accuracy of this model was evaluated; Based on Flex3.0 application program of Adobe Company and ArcGIS Server, the snow cover monitoring information system of the Qinghai-Tibet Plateau was developed by combining ArcGIS theory. The results showed that:1) When the snow depth is above 3 cm, snow classification accuracies of the daily snow cover products MOD10A1 and MYD10A1 are 80.81% and 71.75%, respectively. However, since the serious effection of clouds and other weather conditions, the MOD10A1 and MYD10A1 data almost can not be used to monitor snow disasters in the pastoral areas in practice.2) The user-defined daily,5-day and 10-day combined MODIS snow cover images can eliminate part of clouds, and improve the classification accuracy of MODIS snow cover images. In clear sky conditions, when snow depth is above 3 cm, snow classification accuracies of these three kinds of images are 82.95%,84.78% and 88.78%, respectively.3) The snow classification accuracies of images combined by the MODIS and AMSR-E increased with the increase of snow depth. When the snow depth is above 3 cm, the snow classification accuracies of the daily,5-day and 10-day composite image are 80.48%, 90.83% and 95.95%, respectively. Therefore, the daily composite image is suitable for the Tibetan Plateau snow cover monitoring, multi-day composite image can more accurately reflect changes of the snow cover in the study area.4) Compared with the high-resolution Landsat-ETM images, the MODIS snow cover products underestimated the snow cover area of this study area. The appropriate NDSI threshold of QTP should be 0.35.5) The impact factors of snow depth model on QTP are tempereture, melted snow, rainfall, wet snow, large scale water, deep frost layer, etc. After analysis we find that horizontal polarization brightness temperature data of 18 GHz and 36 GHz bands have a good linear correlation with SD (>3 cm) values, the equation is SD=0.30(Tb18H-Tb36H)+3.18. And 6) we designed the snow cover monitoring information system of QTP, initially completed the research of some snow monitoring functional modules and accomplished the construction task of snow cover remote sensing dynamic monitoring system.