| Qinghai-Tibet Engineering Corridor (QTEC) is located in the hinterland of theQinghai Tibet Plateau, where the eco-geological environment is fragile. Embankmentconstruction, transmission station project, tunnel construction and the left abandonedplaces will inevitablly destroy the eco-geological environment near the QTEC, and dueto the characteristics of particularity, sensitivity and fragility for tibetan environment,the influence of railroad construction on tibetan natural ecological environment may befar beyond the construction itself. Therefore, it is essential to understand the impact ofthe projects along QTEC on the surrounding ecological environment, which hasimortant scientific meaning in strengthing the environment protection and takingpreventive measures.With the development of remote sensing technology, it has more obviousadvantage in measuring surface information. With the increasing diverse needs ofdesertification monitoring and other thematic information, people could monitor thechanges of desertification timely and accurately by the remote sensing technology. Itcould provide the basic data for eco-geological protection along QTEC, strengthen theecological construction and ruduce the ecological deterioration.In this paper,three timely TM/ETM+remotely sensed images including from theyear of2000,2007and2011wre used. Firstly, the pre-processings were taken to theoriginal images, namely, geometric and radiometric correction were done to ensure allthe images were from the same geological referrence and phenology frame. Meanwhile,the geometric and radiometric distortions can also be eliminated in the process.Secondly, the index information of desertification including vegetation and soilabundance were extracted based on spectral mixture models. Linear spectral mixturemodel and gaussian mixture model were used for optimization, and the results showedthat gaussian mixture model has higher accuracy than the linear spectral mixture model.Then the desertification along QTEC was got according to desertification degree indexresulting from sandy gravel ratio. Finally, the Change vector analysis was done to getthe results of changes about the desertification land in the study area. The CVA model could obtain the phase changes of direction and intensity information of the two times.Using the direction information can determine the type of change and the intensityinformation to classify the degree of change by the threshold.Results showed that the total area of desertification has the most obvious changesin5km away from the railway track. Construction has a great interference to thesurrounding vegetation and the distance is mainly concentrated within the range of5km.Moreover, some efforts has also been done to extract the information of sand sourceareas and to get the distribution of sand sources. It could be found from the results thatthere is the main distribution area of sand from Cuona lake to Golmud city, and the mostserious area is near Cuona lake. |