Retrieval Of Vegetation Coverage Using Multi-sensor Remote Sensing Data

Fractional vegetation cover (FVC) is one of the comprehensive indicators to reflect the vegetation coverage on ground, and in the study of ecological system, it is very important. The previous methods of measuring vegetation coverage only can derive the information of a small area, and are hard to retrieve vegetation coverage in a large-scale area. Thus, the use of remote sense technology is necessary and effective in the study of vegetation dynamics. In the study, the retrieval model of fractional vegetation coverage was examined first, and then used to derive FVC information of the study area using the Landsat TM and MODIS data sets. The driven factors of the vegetation dynamics in the study area was also analyzed using some ancillary data.The vegetation coverage information was collected first using the digital camera and the photographs were processed in the ERDAS software application. The in-site FVC data were then used to derive the parameters of the inversion model, and some of the sample data points were also used to validate the model results. In the classification of the digital camera photos, more accurate FVC of vegetation plots were obtained in the crop land, high coverage grasslan, middle coverage grassland, and a lower results were got in the sparsely covered grassland due to the difficulty of plot design.While using different resolution remote sensing data to build inversion models, the ground measured data which will lead to the FVC results with different accuracy. Comparing the two kinds of remote sensing data and two inversion methods, the accuracy of inversion FVC method using TM data is higher than the one using MODIS data, and the lowest accuracy method is the dimidiate pixel model one using MODIS data. To monitor the vegetation cover dynamic change of the study area with the regression model using MODIS data, it can not only fit the feasibility of the inversion, but also discuss the effect between the FVC and the ecologic environment system in big scale.Using the inversion modle, we analyzed the changes of FVC from the year 2001 to 2009.It shows that the general tendency of the FVC area is from the low coverage to the high coverage, especially in 2007 and 2008 with a significantly increase of the high coverage area. This suggests that the vegetation of the study area is in good condition which is helpful to the sound progress of the whole ecological environment. The main natural factors of the vegetation cover changes are the changes of land use and cover type, the amount of the annual regional precipitation, and the level of the annual average temperature. The increase of the high coverage area is closely related to the expansion of cultivated land while the changes of the low coverage area are related to the precipitation.