Abstract: The broadleaved forest Biomass model for the southern parst of Daxinganling Mountains was studied by RS technology in this paper.The SPOT-5 data and Landsat 5 TM in August 2009 were used as source data,with other data like DEM.On the basis of multi-source information composite processing method,normalized difference vegetation index and ratio vegetation index were extracted in ENVI .Slope,aspect and altitude were calculated from DEM.The local plot measured biomass data was taken as standard,through multiple regression analysis,remote sensing model for broadleaved forest Biomass 公式 was established.After accuracy test,all statistics were within the rational range,The average relative error was less than 17.4% showing the predicated values of RS model ere accurate,Less than 17.4%.It can be used for broadleaved forest biomass estimation in Saihanwula National Nature Reserve.Meanwhile,the RS model had a certain practical significance for broadleaved forest biomass estimation in broadleaved forest in the southern Daxinganling Mountains.

Key words: forest biomass, RS technology, multiple regression analysis