Study On Variation Of Grassland Vegetation Based On Multi-Sensor Remote Sensing Data

The grassland is the largest land ecosystem and plays an important role in human being livelihood and development in the process of energy flow and substance circulation. A lot of scientists at home and abroad initiated research on grassland in the early 20th century. As a result of technological limitation, study on grassland just focused on all the aspects of grassland vegetation from the early to the middle of 20^th century, meanwhile research materials mainly came from manual field survey. With the development of research and the advent of new technology, especially remote sensing technology, study concerned made a dramatically rapid progress and it was also possible to monitor grassland vegetation at large spatial level such as national level or global level for a short period of time. Especially in recent decade, remote sensing technology was more and more deeply involved in grassland research. The hot issue in grassland research prioritized the application of remote sensing vegetation indices in the monitoring of grassland vegetation change, and moreover grassland remote sensing gradually became a new discipline.Temperate Steppe located in the hinterland of the Eurasia is a core part of China's northern grassland ecosystem. Because it dominantly distributes in the territory of the Inner Mongolia Autonomous Region, the Inner Mongolia grassland is crucial to the northern ecological environment. Due to green barrier function for the north of China, the Inner Mongolia grassland plays an irreplaceable role in ecological environmental protection and improvement for north China and even the whole country. However, the ecological environment of the Inner Mongolia was severely destroyed day by day from reform and opening up to present. Subsequently, so many environmental problems for instance, grassland degradation, sandstorm and etc., have already seriously influenced the healthy development of national economy, and have directly increased the stress of environmental aggravation of Beijing and Tianjin areas. Therefore, the study on change of the Inner Mongolia grassland vegetation is very significant in improving the ecological environment and promoting the sustainable development of national economy in China.The thesis deeply analyzed selection of remote sensing vegetation indices and models for grassland monitoring, harmonization of TM/ETM+, MODIS, and NOAA data and effect extent and trend of human being actions on the grassland degradation based on multi- sources. Moreover, the mention-above methods were improved and applied to study the two representative regions and the whole of the Inner Mongolian grassland. The thesis also provided a full-picture and integrated analysis for change of grassland vegetation situation of the Inner Mongolian grassland during time period 1982-2006. The study procedure was shown as follows:1) The optimum vegetation index and monitoring model of grassland vegetation situation were chosen targeting at multi-sources remote sensing data and ground data at different sampling scales in the different study areas so as to make them have the value of extension.2) Study on data harmonization of TM/ETM+, MODIS, and NOAA were implemented respectively so as to make the scientific analysis among three kinds of remote sensing data scientific and valid. In addition, with the data harmonization, imageries with high spatial and temporal resolution for regional grassland monitoring were applied to verify monitoring results derived from imageries with low spatial and temporal resolution for grassland monitoring at larger levels. And accuracy verification relation between two kinds of imageries could be further confirmed so as to realize temporal up-scaling.3) Based on the regression model between climate factors and remote sensing vegetation index, study on valid separation of human factor from other factors which are drives for grassland degradation and on effect extent and trend of human factor on grassland degradation were carried on.4) Based on optimum monitoring index, monitoring model and data harmonization model mentioned above, the NOAA and MODIS dada were utilized to study the change of the grassland degradation of the Inner Mongolian grassland for the time period 1982 - 2006, and to analyze the process and trend of that change at macro scale.5) Besides study in (4), the selected representative sampling areas such as typical steppe and meadow steppe at middle spatial level were monitored by using high spatial resolution imageries of TM/ETM+ so as to reflect degenerative process and trend of the Inner Mongolian grassland at different spatial scales and make the monitoring result more scientific, comprehensive and integrated.The main conclusion of this thesis is as follows:1) In the Inner Mongolian grassland regions, R² of the linear model, polynomial model, logarithm model, power function model and exponential model for relationship between ground vegetation situation and remote sensing vegetation index were always high. However, precision of the polynomial model would decline a lot after the verification of ground sampling data.2) In the sampling areas of typical steppe, R² of the regression models between ground vegetation situation data and NDVI, RVI respectively were high; and the coefficient of the regression models between ground vegetation situation data and NDVI, SAVI, RVI respectively for the whole Inner Mongolian grassland were also high. However, precision of the regression model between RVI and grassland vegetation situation would decline a lot after the verification of ground sampling data.3) The strong correlation appeared not only between TM/ETM+ NDVI and MODIS NDVI data, but also between MODIS NDVI and NOAA NDVI data. The correlation turned on stable change regularity over time.4) Data harmonization models between TM NDVI and MODIS NDVI data, and between MODIS NDVI and NOAA NDVI data not only realized the purpose of data harmonization, but also improved accuracy of NOAA NDVI and MODIS NDVI monitoring grassland vegetation situation respectively.5) On the basis of the correlation between the climate factors (the temperature, precipitation) and the remote sensing vegetation indices, the effect extent of human factor on grassland degradation could be effectively separated from that of other factors with better precision.6) The Inner Mongolian grassland degraded seriously during the time period 1982-2006. Compared with 1982, there are only 5 years in which the grassland didn't degrade during time period 1983-2006, the average NDVI vegetation index for the year 2006 was still 0.074431 lower than that of 1982.7) Compared with 1982, degradation situation of the Inner Mongolian grassland turned to be stable after 1998. The vegetation situation of the Inner Mongolian grassland during time period 1999-2006 was inferior to that in 1982 (average NDVI was lower than that in 1982), and there was not even one year in which the grassland remarkably improved.8)Human factor is one of the most important factors leading to the grassland degradation.