| Regional forest ecosystem carbon’s reserve and change have a huge impact on global carbon balance. Whatever the present and the future, the research on climate change and carbon balance are of great significance. For the study of land system’s carbon cycle and global carbon cycle, forest ecosystem carbon is a foundation.The study base on the Xiaoxing’an mountains area’s remote sensing image and 130 pieces of sample plot survey data, set up two groups of carbon density model. First, each stand factors as independent variables and choose them. Base on conventional stepwise regression method to establish forest ecological system or forest ecosystem carbon density model of the total and the compatibility forest ecosystem carbon density model of the field investigation. Then compare to the precision of two models.Secondly, preliminary choose the grey value of all the bands, the different linear and nonlinear band combination between the grey value(including all kinds of vegetation index), texture information, the auxiliary band after abiotic factors rasterize---171 independent variable of environmental factors to test correlation coefficient significance of tje carbon intensity. The estimation of forest ecosystem carbon density of Xiaoxing’an mountains, carbon density including vegetation carbon density (the ground vegetation and root) of carbon density, soil carbon density and litter carbon density, then precision evaluation. The main conclusion are as follows:1) In each stand factor and total carbon density estimation model, for compatibility forest ecosystem carbon density model, fittings results of the leaves trees, roots, trunk, the earth trees and vegetation carbon density model are good, while the fitting results of shrubs herb and soil carbon density are not good as expected. But compared with stepwise regression method, compatibility forest ecosystem carbon density model still estimate each component and the total to improve, with nearly 7% of the increased significantly.2) Simultaneous equations model has good application in forest stand growth model, but the model is less applied to extract carbon density on remote sensing. This study with the simultaneous equations model method will be introduced to estimate remote sensing’s forest ecosystem carbon density. On the one hand, the method explore a new remote sensing estimation model, on the other hand ,it provides a new way to estimate multisensor project information.3) In remote sensing variables and the total sample area carbon density estimation model, compare the normal statistical model(4-4) introduction of crown density variables with the normal statistical model(4-3). Due to the accuracy of the field investigation crown density or of the error recorded is not significant improved, crown density model estimatation don’t meet the accuracy requirement, so crown density fuction of the variables is smaller in the regression model type. However, after carbon density estimation model and canopy density estimation model simultaneous equation parameters is established, can effectively reduce the error of the two equations. The final carbon density estimation model type (4-6) precision is improved. Studies have shown that simultaneous equations model for estimating forest ecosystem carbon density is the optimal model and the accuracy is improved by 6%-7%. This is a new thought for estimating forest ecosystem carbon density on the remote sensing.4) The south and central parts of Xiaoxing’an mountains, the forest carbon density is low, mainly concentrated in 200-250 t-hm-2. The south of Xiaoxing’an mountains, the forest carbon density mainly concentrated in 250-300 t-hm"2. The forest ecosystem carbon density in Xiaoxing’an mountains from west to east and from north to south has a gradually rising trend.
|
PDF Full Text Request