Abstract: A data-driven systematic error estimation method for radar system is proposed to solve the problem of poor systematic error correction e ect in complex physical environment. A numerical model is used to divide the radar detection area into grids. After the error distribution function is established on the grid, the variance iterative calculation is used to estimate the parameters of error distribution function on each grid. The proposed method overcomes the diculty of establishing the radar systematic error model with complex parameters and obtains the distribution function of the systematic error by numerical method. The simulation results show that the proposed method has a better accuracy of systematic error estimation with a small calculation and simple engineering implementation, when there are automatic identification system (AIS) and other high precision information sources.