| This thesis was based on the experiments carried out on three samples with different land forms of slope, plain and hilly in the west Ujimqin Steppe from 2012 to 2014. The data of vegetation coverage, species, average height, species coverage, species average height, density, biomass, chestnut soil thickness, soil mechanical composition, soil moisture of 150 quads per year was collected from each sample, the inter-correlation and significance between data types was analyzed, vegetation coverage, thickness of chestnut soil and value of Leymus chinensis+Stipa grandis was chose for building the steppe degradation index based on the variation of chestnut soil for the purpose of evaluating steppe degradation synthetically. Combined three years remote sensing image of Landsat satellite, the invertion method of steppe degradation index was researched, the model of vast scale monitoring of steppe degradation by using remote sensing was built, the degradation situation of three samples in 2012 to 2014 were monitored and analyzed, the effect of monthly precipitation on steppe degradation was determined primarily, the accuracy of the model of monitoring steppe degradation was proved by the error analysis of steppe degradation index of invertion and field survey. At last, this thesis recorded the study of desertification area and its neighborhood of the trial in typical steppe, extrapolated the approximate range of threshold of typical steppe desertification, built the evaluation model of typical steppe desertification risk based on the threshold, evaluated the desertification risk of three samples, set up a series scientific, reasonable and practical methodologies of steppe monitoring and evaluation, provided the scientific basis of steppe degradation and desertification control and ecological assessment. The main conclusions of this thesis are as follows.1. The statistics and analysis result of the data of vegetation coverage, species, average height, species coverage, species average height, density, biomass, chestnut soil thickness, soil mechanical composition and soil moisture showed:Vegetation coverage, average height had no direct relationship to the land form, but there was significant difference of vegetation species distribution between different land forms; The chestnut soil thickness of three different samples in three years had no obvious change; The fixation effect of surface soil moisture of vegetation in medium flow year is more prominent than in high flow year.2. Based on the research result of retrogressive succession of vegetation species, the value of Leymus chinensis+Stipa grandis that expressed the homoplasy between typical steppe vegetation and its climax or sub-climax state was built with the species coverage and average height of Leymus chinensis and Stipa grandis. The point-biserial correlation between the value of Leymus chinensis+Stipa grandis and the discrimination results of dominant species showed that the Pearson correlation coefficient is 0.536,0.538 and 0.451, and the correlation pass the test of significance, it means there was medium significantly positive correlation between them, the validity of the value of Leymus chinensis+Stipa grandis had been verified.3. Through the correlation analysis of vegetation, soil data, filtered the vegetation coverage, chestnut soil thickness and the value of Leymus chinensis+Stipa grandis with low correlation between each other to built the model of steppe degradation index, the synthetic model of three samples are:the model of steppe degradation index of 2012 is SDI= 0.31C’+0.27 LSV’+0.4A2T’, the two years model of steppe degradation index of 2012-2013 is SDI= 0.32C’+0.28LSV’+0.4T’, the three years steppe degradation index model of 2012-2014 is SDI= 0.30C’+0.27LSV’+0.43T’. And based on the three years model of steppe degradation index, plotted the distribution map of steppe degradation index of different sampling gradients in three samples, analyzed and synthetically evaluated the steppe degradation situation of different sampling gradients of different land form samples in high flow year 2012. The evaluation result showed that the mild to severe order of three samples were:Slope sample> plain sample> hilly sample. But the restoration succession potential of plain sample was lower than slope and hilly samples, because of the thickness of chestnut soil of plain sample was thinner.4. After the preprocessing of destriping strip, radiation calibration, band synthesis, atmosphere correction, geometric correction of the remote sensing image 2012-2014, obtained the remote sensing parameters of each band image and several vegetation index images to build the 2012 invertion model of steppe degradation index vegetation part with difference vegetation index DVI based on the result of correlation and regression analysis between the steppe degradation index vegetation part of field survey in 2012 and the remote sensing parameters. And built the 2013 and 2014 invertion model of steppe degradation index vegetation part by using radiation registration between the DVI of 2012 and the DVI of 2013 and 2014, the error analysis showed the invertion models accuracy of 2012-2014 is 83.85%,84.68%,83.31%, which meets the requirement of the synthetic evaluation of steppe degradation on vast scale. Then the invertion result of vegetation part of three years steppe degradation index model and the Kriging estimate result of chestnut soil thickness was taken as the factors, the three years steppe degradation index model was taken as the basis, the steppe degradation remote sensing monitoring model based on the variation of chestnut soil thickness 2012-2014 was built:The 2012 model of steppe degradation remote sensing monitoring was SDI2012=2.137DVI2012+0.43T2012-0.201, the 2013 model was SDI2013=1.120DVI2013+0.43T2013-0.030, the 2014 model of steppe degradation remote sensing monitoring was SDI2014=1.214DVI2014+0.43T2014+0.019.5. Based on the field survey data 2012-2014, the cluster analysis of vegetation coverage, thickness of chestnut soil, steppe degradation index of typical steppe trial was conducted, the steppe degradation grading standard of the factors of vegetation coverage, chestnut soil thickness and steppe degradation index was formulated. Based on these grading standard the vast scale monitoring of the trial steppe degradation situation was-carried out. The monitoring result showed:The steppe degradation of slope sample in 2012 was mainly middle degradation, the middle degradation area was 75.6% of the slope sample area, severe degradation area was 11%. The steppe degradation of plain sample and hilly sample in 2012 was mainly severe degradation, the area was 42.6% and 54.4% of each sample area. The situation of steppe degradation in 2013 and 2014 were almost same, the middle degradation area of the slope sample in 2013 dropped to 69.1% of the slope sample area, the severe degradation area of slope sample in 2013 rose to 30.8% of the slope sample area, the severe degradation area of the plain sample and hilly sample in 2013 rose to 63.1% and 79.7% of each sample area.Through the correlation and regression analysis, the steppe degradation index model of monthly precipitation was built: 100SDI=0.506PreMar+0.026PreJun+8.493. This model express the causal relationship of that the general level of typical steppe degradation was strongly influenced by the monthly precipitation of March and June.6. Through the statistics and analysis of the vegetation and soil data of desertification area and its 50m,100m neighborhood, extrapolated the approximate range of threshold of typical steppe desertification, built the synthetic evaluation model of typical steppe desertification risk that is composed of the membership function and fuzzy evaluation matrix of typical steppe desertification risk based on the threshold and fuzzy mathematics theory. And the desertification risk of three samples in median flow year 2013 were evaluated, the risk evaluation result of typical steppe desertification was:The slope sample was B1= W。R1= [0.4478,0.5502], the plain sample was B2= W。R2= [0.4906,0.5094], the hilly sample was B3= W。R3= [0.4547,0.5453]. It showed that the desertification risk of three samples still belonged to general risk, but the value of evaluation was very closed to the high risk, especially the plain sample. If reasonable steppe ecological protection measures still can not implement on the plain sample, when the low flow year comes, the steppe ecology will be worse, the area of desertification will rapidly expand, the degree of desertification will gradually increase, the steppe ecology balance will be broken, the function of steppe ecology system will disappear. |
PDF Full Text Request