| In this paper, remote sensing inversion model of forest biomass in Beijing was established based on3S technology, and TM remote sensing data was used for the data source. The vegetation of the remote sensing data can be divided into broad-leaved forest and coniferous forest. Combined with field plots data,20modeling factors (including vegetation index factor, band combination factor and topographical factors and so on.) was selected to set up respectively biomass inversion model of broad-leaved and coniferous forest based on the relationship between forest biomass and the vegetation index and terrain factor. The methods included principal component linear regression, multiple linear regression, the BP neural network, etc.Through the establishment of multiple models and analysis, providing a precise basis for further analysis of remote sensing for precision forest biomass model.Through component analysis, correlation analysis and other methods, the relationship between each factors and biomass was got. And in according to its relationship, the modeling factors was selected based on their correlation. Through modeling study of three methods, the result showed that the accuracy of the model established by principal component regression method is relatively low, and the neural network model has the highest accuracy. Prediction accuracy of the three methods for broadleaf forest are80.57%,82.2%,89.5%respectively, and the coniferous forest are79.63%,83.9%,92.5%respectively. Compared with the advantages and disadvantages of each model, select the most appropriate model to make an estimate.The estimation result is that the biomass of broadleaf forest is46,403,664.3tons, coniferous forest is18,229,811.8tons, the total biomass is64,633,476.1tons. And the forest biomass was divided five levels, and output biomass level maps. Biomass distribution of â…¡, â…¢ grade is the most relatively, the northern mountains are areas that higher biomass mainly distribute. After estimate biomass of Beijing, we statistics biomass of16districts and counties in Beijing. The result showed that Huairou has the largest forest biomass, the followed are Miyun, Yanqing, Mentougou. |