Satellite Remote Sensing of snow cover over Northeast China

Climate change becomes the hottest focus in the world at present as the tremendous increases of frequency of extreme weathers and natural disasters. It is not the problem only for government or scientists. Civilians start to concern about this issue when flooding, snow disasters and earthquakes are happening all over the world. The need to understand what is going, what will happen and why it happens on global climate becomes imperative. The study on snow cover provides us a possibility to fulfill the need.;Changes of snow cover is not only a result of global climate change, but also a contribute factor to the changes. The distribution and amount of snowfall play critical role in energy balance, heat release and atmosphere circulation. The changes on snow cover wi11lead to local, even global changes of climate, not to mention local disasters.;In this thesis, a systematic study on snow cover over Northeast China (Heilongjiang, Jilin and Liaoning provinces) is conducted. Firstly, study on the properties of local snow has been done by the method of insitu measurement and ground based field experiments. The result shows that, as one of the most sensitive parameters, snow grain size varies from 0.2mm to over 3mm. The range of snow grain size is too big to determine a fixed constant in passive microwave snow monitoring models. The followed experiments result on snow grain size and reflectance suggests that the problem can be solved by using a timely snow grain size retrieval model based on NIR reflectance data. Then the automatic process methods for mass of MODIS and AMSR-E data are discussed to get high quality input for snow covered area and snow depth model. By using the new process methods, the processing time of a 48 working days' job can be finished within two working days. After the preparation works, two improved models for snow covered area and snow depth monitoring are developed. To compare with MODIS snow products, the SCA monitoring accuracy of improved model is better. Less errors on snow covered test sites can be found in improved model in the case study on SCA monitoring over Northeast China.;An initial data assimilation conception is introduced into SD estimation model. In improved model, snow grain size increase ratio that calculated by using MODIS band 4 and band 5 reflectance data are inputted into passive microwave SD estimation model to determine the value of constant which is changing according to the changes of snow grain size. Results show that, during the dry snow period, in new model the RMSE of estimation SD and measurement ones in 5 test sites are increased 24%, 0.4%, 23%, 65% and 16% respectively. This data assimilation method provides a bright new way in combination application of optical and microwave remote sensing technology in snow monitoring.;Although many achievements are obtained in the research, the investigation on snow properties and the preliminary application of remote sensing on snow monitoring during the 4 years from 2005-2009 shows us there are still some problems that need to be solved, they are: (1) Snow depth monitoring during wet snow period; (2) Snow depth monitoring on ice; (3) Specific snow grain size estimation by using MODIS data. The future work on snow monitoring should go deeper for better estimation accuracy.