Abstract: The influence of ground scene sampling-rate and sensor scanning direction has been examined to accurately simulate the sub-pixel line target in synthetic spectral imagery. Both theoretical analysis and simulation experiments were proposed. The geometric relation between the pixels of the satellite detector and the ground scene was inferred. Hough line transform was adopted to describe the radiometric property and spatial distributions of the simulated line target. Results were analyzed with the DataFrame. It is concluded that the aliasing error increases notably with the angle between the scanning direction and line varying in the range near 0, 90. The line becomes less significant and fake lines appear. The similar phenomenon exists in the angle range near 45. However, the aliasing effect is relatively mild, thus providing robustness to the simulation. To a certain extent, sampling with a more intensive rate improves the line's significance, but cannot eliminate the blur-introduced errors of spatial location. These results can be used to evaluate the simulation fidelity, and design appropriate models.