Abstract: Shallow subsurface investigations utilizing microtremor inversion for shear-wave velocity (Vs) face various uncertainties in delineating soil layer structures. It is essential to analyze and correct data to enhance the practicability of Vs data adopted in soil stratigraphy identification. Previous studies on Vs-depth relationships exhibit obvious regional limitations and are inappropriate for direct application in the Black Soil Region of Northeast China. This study employed microtremor inversion results and borehole data from the northern part of Songnen Basin. After comparing the applicability and accuracy of five regression models for describing the Vs-depth relationship, we identified the quadratic function model as the optimal. Subsequently, we established quadratic regression equations relating Vs to depth for 11 major soil types within the study area, enabling the determination of depth-specific correction values for each soil type. Validation was performed by using two case studies with loam as the reference soil. Results demonstrate that the quadratic function model effectively captures the non-linear relationship between Vs and depth for these soils. After correction, the Vs profiles showed high coincidence with the loam distribution revealed through boreholes. Depth correction substantially mitigates the soil compaction effects induced by burial depth. The corrected Vs values significantly improve the accuracy of soil layer structure identification.