Based On High Resolution Image The Green Biomass Research In Hangzhou Xihu District

City is the most concentrated area of production and life of mankind. Studies have shown that 97% of man-made CO2 come from urban areas. Urban green space ecology system is an important part of terrestrial ecosystem. It has an irreplaceable role in the process when absorbs CO2 and releases O2. Therefore, the study of urban green space biomass accurately estimated is of great significance. The study selects the Hangzhou Xihu District as the target, combining field sample data with remote sensing data. Then choosing the variable factors to do correlation analysis. The better independent variable factor correlation would be selected to build remote sensing estimation model of green biomass. Next, to test the fitting precision of the regression model by the reserved experimental data. Choosing the model which has high precision and better estimation for the green biomass inversion in the study area. The results mainly includes the following aspects: 1. Based on the SPOT6 high resolution images, the study uses object-oriented image analysis to extract green space information in different blocks. The vegetation coverage of forest area is the best( 92%), while the Suburban farmland is the worst(only 27%). 2. Combing the fifty-two independent factors associated with green biomass with the biomass of field samples to do the correlation analysis. As a result, there are nine variables factor is better, such as Band3, NDVI, RVI, NVI, Mean2, Variance3, Entropy3, Dissimilarity4 and PC1. Among them, Band3, NDVI, RVI, Variance3 and Entropy3 have the highest correlation, which is more than 0.5. 3.The study selects the better independent variable factor to build the biomass estimation model by different methods. The results show that the stepwise regression model is better than the multiple linear regression model, which has better fitting effect and better test accuracy. The fitting precision is 77.24%, the test accuracy is 74.71%. Therefore, the study selects stepwise regression model to estimate the green biomass in the Hangzhou Xihu District. 4. About the urban green space biomass analysis.The average biomass is 34.75t/hm2, the total biomass is608259.23 t. The average biomass of mountain forest is in a highest scope, which is more than 45.00t/hm2. The suburban and urban areas are the lowest, which are mainly at 0-15.00 t/hm2 and 15.00-25.00 t/hm2.