Abstract: A remote sensing driven dynamic simulation model was developed for terrestrial ecosystems. The model was encoded in C language under the environment of SPAMOD, a spatial simulation tool developed under MS Windows. The model was applied to Northeast China Transect to simulate the dynamics of green and non-green biomass of 12 vegetation categories as well as soil water of 3 layers. The green biomass was converted to normalized difference vegetation index (NDVI) of AVHRR remote sensing, and compared with the observed NDVI from 1986 to 1990. The model was also compared with ground measurements of biomass and productivity along the transect. Ambient CO2 concentration, monthly mean air temperature and monthly precipitation were regarded as the three basic driving variables for global change study. The model also included the effects of temperature and precipitation on sunshine fracti6n, relative humidity, radiation, soil water and eventually plant growth. For each CO2 and climatic scenario, the model was run for an equilibrium solution. The results indicated that the natural vegetation of the transect was very sensitive to variation of temperature and CO2 concentration. With CO2 remained unchanged and temperature increased by 4 CE, the induced increase in evapotranspiration could reduce the average biomass and net primary productivity (NPP) over the whole transect by 32.1% and 41.9 % respectively. In contrast, a 20 % increase in precipitation alone could lead to an increase of the average biomass and NPP by 8.1% and 13.4% respectively. Under the present climatic conditions, CO2 doubling could increase the average biomass and NPP by 12.2% and 17.1% respectively. Because of compensation between the positive effects of CO2 and precipitation increase and the negative effect of temperature increase, a comprehensive interaction among CO2 doubling, a 20% increase of precipitation and a 4 ℃ increase of temperature altogether can lead to approximately a 2% reduction in the biomass and NPP of the natural vegetation over the whole transect.

Key words: Ecological transect, Vegetation dynamics, Remote sensing, Global change