Seismic decomposition discriminates different geological expressions by isolating seismic signals of particular frequency ranges. As a novel signal decomposition method, Variational mode decomposition (VMD) exhibits advanced features compared with the classic Empirical Mode Decomposition (EMD) method. VMD can resolve the instinct mode functions (IMFs) more robust and reasonably. Besides a synthetic example, we show a field application of the VMD in sedimentary cycle identification. However, like other high resolution decomposition method, traditional VMD also decomposes data trace by trace, which would result in poor lateral consistency. In order to appropriately apply it on seismic data, a lateral consistency preserved VMD method is proposed to meet the challenges of seismic application, which has been verified through field applications to be valuable for further analyses.