Abstract: The PSInSAR technique can achieve separation of PS phase components, by extracting the time-dimensional PS points based on various temporal and spatial statistical properties to get high-precision, surface deformation monitoring results. As the main error source in InSAR, atmospheric signals can be isolated from the other components of the residual phase by classical filters in the spatial and temporal domains. Optionally, after 3D unwrapping, high-pass filtering can be applied to unwrapped data in time followed by a low-pass filter in space in order to remove the remaining spatial correlated errors (atmosphere and orbit errors). Thus, when the deformation rate is large, the spectrum of various contributing factors will overlap, and the classic filter is powerless. This paper proposes methodologies which can automatically choose a smoothing parameter based on a fast robust version of a discrete smoothing spline instead of classic filter to effectively separate the phase components.