Abstract: It has been found that some forest biological parameters suchas the average tree height，biomass，and volume can be effectively estimated from Synthetic Aperture Radar(SAR) remote sensing data. However，the data accuracy is affected by the uncertainties in SAR systems，as a result，the accuracy of forest parameter inversion could be lost or even abnormal. simulation of SAR data as three dimensional forest scenes permits uncertainties to be decoupled to generate mass of data with specified parameters，which can be used for analyzing the uncertainties separately or together. SAR three-dimensional forest simulation models considering uncertainties were established and verified by E-SAR measured parameters and data. Supported by a large amount of simulated data，the uncertainties in a SAR system are quantitatively analyzed for residual phase errors after motion compensation. The relationships between residual phase errors and RMSE of forest height inversion were obtained.