Since grassland ecosystem is one of the most widely distributed terrestrial ecosystems, the accurate statistic of the biomass of the grassland ecosystem is a key issue in terrestrial ecological studies. Presently, the primary methods to compute the biomass of the grassland are almost based on various vegetation indexes, of which the NDVI is employed most frequently. However, the calculation of vegetation indexes contains many uncertainties, which can be brought into the grassland biomass estimation. In order to avoid these uncertainties in the vegetation indexes, this paper tries to build a new model to compute grassland biomass which is independent of various vegetation indexes. Firstly, this research divides the study area into four kinds of potential vegetation types according to the annual cumulative temperature (> 0℃) and the K -value (humidity index) based on Integrated Orderly Classification System of Grassland (IOCSG). The first kind of the potential vegetation type is Cool temperate-arid temperate zonal semi-desert, the second kind is Cool temperate-semiarid temperate typical steppe, the third is Cool temperate-subhumid meadow steppe and the last one is Cool temperate-humid forest steppe, deciduous broad leaved forest. Then, correlation analysis was applied between aboveground biomass of the grassland and some selected impacts for each potential vegetation type. The criteria of the impacts selection aim to make sure that the factor surely has some impacts on the biomass estimation, but not the one that reflects the biomass. For example, both the temperature and the rainfall belong to the former, while the NDVI belongs to the later. In this paper, annual cumulative temperature, the K -value, potential NPP, slope, aspect, clay content of the surface layer and sand content of the surface layer are selected, of which the potential NPP is calculated according to annual cumulative temperature and the K -value. The analysis results show that the main impact of the aboveground biomass of the grassland on each potential vegetation type is different. For Cool temperate-arid temperate zonal semi-desert, the main impact is annual cumulative temperature; On the part of Cool temperate-semiarid temperate typical steppe, the main impact is the clay content of the surface layer; and to Cool temperate-subhumid meadow steppe and Cool temperate-humid forest steppe, deciduous broad leaved forest, the main impact is potential NPP. However, the relationship between the biomass and the potential NPP is different for the above two potential vegetation types. Since the main impacts of the grassland biomass for each potential vegetation type is confirmed, mathematical relationship is built respectively. Finally, High Accuracy Surface Model (HASM) was employed to simulate the biomass of the grassland in the study area. Simulation results show that the aboveground biomass density of the grassland in the four potential vegetation types is 76.62, 110.94, 142.69 g/m² and 184.40 g/m² respectively.