| This dissertation proposed vegetation quantity index that was capable of describing the 3-dimensional characteristics and dynamics of vegetation, with consideration of chromaticity of greenness and accurate estimation of vegetation quantity, and at the same time applicable for remote sensing data. Based on the remote sensing imagery and synchronized measurements from sample plots, empirical retrieval model of vegetation quantity was built, and effort was made on building practical and accurate retrieval vegetation quantity models based on pixel decomposition. Using the built retrieval models and related remote sensing imagery in the research area, a variety of forest types, slope aspects, and forest vegetation quantities at different altitudes in mountain area of southwestern Beijing were identified, and also analysis the spatial distribution and seasonal dynamics of forest vegetation quantity in mountain area of south-western Beijing. The results were as follows.(1) This study proposed using the product of leaf area index (LAI) and chlorophyll volume index (CVI) to describe the vegetation quantity, i.e. VQ=LAI X CVI. This was essentially an index with consideration of chromaticity of greenness, capable of better describing 3-dimensional characteristics of vegetation with high sensitivity on chlorophyll volume. Application showed that this index can estimate vegetation quantity more accurately from remote sensing imagery.(2) Comparisons among regression models, estimated vegetation quantities, scatter plots of measured values, and monitoring of seasonal forest vegetation dynamics showed that the power equation with TSAVI as the independent variable can be a good estimator of vegetation quantity (R2=0.918. RMSE=1.534). The empirical retrieval vegetation quantity model is VQ=14.892×TSAVI0.779. (3) Based on the validation of prediction accuracy using measured values of leaf area index and retrieved values, it was shown that retrieval vegetation quantity model based on pixel decomposition, which retrieves leaf area index by pixel decomposition, can meet the demand of accuracy. Thus it is useful in large-size vegetation quantity retrieval and dynamic monitoring.(4) There was significant heterogeneity in the spatial distribution of forest vegetation quantity in mountain area of south-western Beijing, with vegetation concentrated between 500m-1500m in altitude. In terms of the pattern of vegetation quantity distribution, forests in the mountain areas between Fangshan and Mentougou as well as between Fangshan and Hebei had high intensity of vegetation quantity. The average vegetation quantity in mountain area of southwestern Beijing was 8.378, with averages for broad-leaved forest, coniferous forest, and shrubbery being 8.028,8.354, and 8.905 respectively. The distribution patterns of vegetation quantity were significantly affected by the types of forest stand. Slope aspect had some effect on vegetation quantity, with averages at shady slope and sunny slope being 9.015 and 7.986 respectively. Shady slope had higher vegetation quantities than sunny slope at all four altitudes with the same type of forest stand. Average forest vegetation quantity increased significantly with altitude, with average vegetation quantities being:6.764 below 500m,8.54 between 500-1000m,9.318 between 1000-1500m, and 11.64 above 1500m. It was obvious that average vegetation quantity below altitude of 500m was low because forests there were vulnerable to destruction and disturbance, with significant difference between shady slope and sunny slope.(5) The dynamics of vegetation quantity in mountain area of southwestern Beijing were highly correlated with climate. The average forest vegetation quantities in different months were: March 1.47,May 3.789, August 7.129, November 1.629, December 0.966; growth season (August) had significantly higher value than other months; the seasonal dynamics of vegetation quantity at shady and sunny slopes can be both described by parabolic curves, with maximum and minimum at August and December respectively. The vegetation quantities at four altitudes all achieved their maximums in growth season (August) and their minimums in stagnation season (December). The vegetation quantity index proposed in this study can better describe the actual vegetation quantity from remote sensing imagery. Using the index proposed by this dissertation, the results from the research on forest vegetation quantity in mountain area of southwestern Beijing based on remote sensing imagery and data measured at sample sites, can provide technical support for the assessment of ecosystems in districts of different functions in Beijing, serving as references for monitoring and assessment of the benefit of forest ecosystems in Beijing.
|
PDF Full Text Request