| It is a unique geographical environment in Xinjiang, which deeply located inland with arid climate and fragile environment. At present, the monitoring of vegetation cover change has mature methods, but without effective interpolating of climate factors and quantitative spatial expressing of human factors. As these reason, it is so hard to explore the spatial correlation between vegetation cover change and climate and human factors.As a natural tie of connecting to soil, atmosphere and water, vegetation is like an indicator in the global change research. Vegetation cover change present significant spatial-temporal difference characters, which is the result of long-term interaction by nature and human activities. Under the situation of human activity effect becoming increasingly important, it is a very important aspect of global change research to study on the relationship between vegetation cover change and climate and human activity. Combining with the long-term remote sensing vegetation data, meteorological station data (temperature and precipitation), annual statistical data and field investigation data, the author discussed next several problems by dividing the whole area into four sub-areas in geography and then taking geomorphologic division by altitude from 1982a to 2006a in Xinjiang area. The first is that we revealed the vegetation cover change spatial-temporal characters through using the methods of trend, regression, correlation and variability analysis. The second problem is how to treat meteorological factors. We simulated the spatial distribution information of precipitation and temperature by using stepwise and integrated methods based on considering multi-factors, analyzed their dynamic tendency. The third is that we selected and extracted many human factors information based on comprehensive survey, given spatial quantitative expression and dynamic analysis by regional separation. The last but most important thing is that we explored the relationship between vegetation cover change and climate, human activities, and explained the reason, especially the human factor and spatial effect of vegetation cover change in Xinjiang by using the methods and models of spatial econometric analysis, from the view of spatial-temporal correlation integrated with climate and human factor change character. Through these works, several conclusions have been gotten.1) The spatial-temporal characters of vegetation cover change in Xinjiang area indicated that the fvc dynamics increased with significance in general, but fluctuated remarkably in period from 1982 to 1995 and relatively stable from 1996 to 2006, varied significantly with subareas, and southern Xinjiang increased more obviously than northern, Yili plain increased apparently, but decreased in mountain, and without significant trend in eastern Xinjaing. The vegetation cover changed remarkable difference in plain and mountain or under different vegetation and landuse type. These indicated distinct zoning feature, vertical difference and diversified controlling factors. For the aspects of fvc change amplitude and variability, the eastern and southern Xinjiang area were greater than northern Xinjiang and Yili area, and the lower fvc was greater than the higher fvc area, the plain area influenced more by human activities was higher than mountain area effected more by natural factors. The time correlation and change rate of fvc variation were consistent with each other in spatial distribution, which positive correlation area located in oasis, oasis surrounding region and part of mountain with positive change rate, but negative correlation region was situated in oasis-desert transition and desert zone with negative change rate. These showed that positive correlation area located in higher fvc area, but negative correlation region was situated in lower fvc area. In order to characterize the fvc dynamics, annual average fvc was divided into four characteristic intervals: 0.00~0.10 called degenerate interval, 0.10~0.25, slight improvement, 0.25~0.90, significant improvement and 0.90~1.00, saturated interval, which show that the smaller of fvc, the more obvious of degeneration, the higher of fvc, the more obvious of improvement tendency, and improvement decreased significantly as fvc becoming saturated. This spatial-temporal difference feature was a result of specialty of mountain-oasis-desert system and fvc response to hydrothermal condition change and long term effect by human activities.2)For the climate change research, we established some simulative meteorological stations as a supplement of spatial variability information of fact station data through finding weather stations influence zone by investigating appropriate sampling interval distance of multi-year average precipitation with geostatistical analysis, and founding multiple regression equation between precipitation and longitude-latitude, elevation, slope, aspect and NDVI(normalized difference vegetation index), which can effectively enhance the precipitation spatial variability information and interpolation precision. Then we used the interpolation methods of Multiquadratic Radial Basis Functions (MRBF) to interpolate precipitation in Xinjiang area, and the annual precipitation spatial distribution dataset of this area from 1982a to 2006a with higher precision had been obtained. For the temperature interpolation, we used two self-constructed mixed methods, the"multiple regression + residual interpolation + topography regulation"and"elevation effect correction + zero elevation interpolation + topography regulation", to simulate the air temperature spatial distribution for multi-year month or year average temperature, at same time, cross validation and comparison analysis had been taken. The annual temperature spatial distribution dataset with high precision from 1982a to 2006a have been gotten by the method of"elevation effect correction + zero elevation interpolation + topography regulation". The precipitation increased significantly from 1982a to 2006a in Xinjiang area. The area of increasing was more than the decreasing area. The region with higher precipitation increased more obviously than the region with lower precipitation, and further decreased in some of the extreme dried region, but most area have not significant change trend in Xinjiang. The temperature was going up obviously from 1982a to 2006a, and mountain raised greater than plain, desert raised quickly than oasis. In conclusion, in 25 years, the climate change was that both temperature and precipitation rose significantly, and with pronounced regional difference.3)With regard to the human factor, based on Xinjiang statistical yearbook, landuse data, atlas, field investigation data, after put forward the principle of indexes extraction, we extracted gross domestic product(GDP), density of population, urbanization rate, grain output, number of year-end livestock, per capita gross domestic product, total meat product, primary industry product(agriculture, forestry, livestock, fishery), boundary of oasis, cities and road, 10 human factors, which can take spatialization, and area of arable land, cultivated, irrigated, 3 can't do spatialization factors. In order to make human factors spatialization reasonably, we constructed feasible spatialization expression rules, and based on these rules, 8 spatializable human factors have been spatialized. Then we got 8 human factors spatial distribution map from 1988a to 2006a, and oasis, cities, road maps. All of these data have been added into the geographical information system (GIS). At last, we explored the time sequence change trend and spatial location characters of human factors from 1988a to 2006a by time series and regression analysis methods under dividing the whole area into 4 sub-regions. Some useful conclusion has been received.4) Based on the above work, we analyzed the factor of influence of vegetation cover change. The result showed that the correlation of precipitation and vegetation cover was greater than temperature. The region of positive correlation between vegetation cover and precipitation was located in eastern and northern Xinjiang, and in southern Xinjiang between vegetation cover and temperature. The dependency and sensitivity of vegetation cover change to precipitation and temperature are different from vegetation and landuse type. The increasing of precipitation may improve the vegetation cover, but the rise of temperature may lead to the degeneration of vegetation. There is obvious difference in spatial location. The region of significant correlation between vegetation cover and precipitation was in higher vegetation cover and precipitation, and in less precipitation, higher temperature and lower vegetation cover between vegetation cover and temperature. Vegetation cover influenced by human activities was increased along with the distance to oasis and road, which has significant different characters. On the whole, the improvement rate descends, and degeneration rate ascend along with the increasing of distance, but improvement rate was always greater than degeneration rate. The correlation between vegetation cover and every human factor were weak, significant in oasis of southern Xinjiang only. Spatial similarity has displayed for different human factor to vegetation cover, that positive significant is in oasis and its surrounding areas, and negative is in part of mountain and desert which are far away from human activity. It is bidirectional for human activity influences that improvement occurs in some area, but indirectly degenerated in other area, which the difference and relation coexisted in spatial position.5) 8 human factors have been treated with dimension reduction by factor analysis, then 4 common factors (development level of agriculture-forestry and grazing; economic and consumption level; oasis cover level; population loading level) have been gotten, that represents the influence degree and range of human activity well. Result showed vegetation cover change character has significant spatial autocorrelation and differentiation. The spatial effect of vegetation cover change character has been analyzed and answered using the analysis of spatial econometric, combining with the statistical tools of GeoDa and SPSS, by the models of classic regression, spatial lag and geographical weighted regression (GWR) under ArcGis 9.3. Climate and human activity are two aspects of effecting vegetation cover change, but both have direct and indirect ways. From quantitative view, the direct effection of climate and human activity account for about 5.4% and 4.6% respectively. But the neighborhood effect (42.5%) and spatial heterogeneity (46.5%) have explained the residual reason that indirect common effected by both factors. Vegetation cover is a long-term dynamic balance process, that changed from seriously fluctuate to relative stable from 1982a to 2006a. The climate change presented as"improving temperature and increasing precipitation"process, but with obvious regional differences. Most of the human factors increased continuously without volatility. The effects to vegetation cover change take in 4 aspects, which are development level of agriculture-forestry and grazing, economic and consumption level, oasis cover level and population loading level, that taking oasis as influencing center, traffic road and river as mainline form the core zone of action, then radiate and penetrate to mountain and desert form the buffer zone of influencing. The effection has direct and indirect two ways, improving in core zone, but negative or no influence in buffer zone and faraway. In this paper, the author has quantitatively answered the influence weight of climate and human factor in vegetation cover change, and the question coming with it about spatial neighborhood effect and spatial heterogeneity. So the vegetation cover change character has been quantitatively explained and analyzed.
|
PDF Full Text Request