Seismic data composite a data carrier of the spatial change relationship,which reflects the characteristics of changes in the geological space.Under the premise of known seismic data,the problem as to how to discover the spatial variation of reservoirs is an important issue in fine reservoir research.Based on the study of the stratigraphic spatial transformation function,this paper proposes a method for obtaining stratigraphic space transformation functions based on seismic trace matching.In this method,multi-channel seismic traces are used to obtain the weighting coefficients,and the multi-point convolution-weighted formation space transformation function is used to replace the point-to-point formation space transformation function,so as to accurately reflect the physical property changes and reduce errors.The actual data of Q oilfield,and the model testing and practical application analysis show that,compared with traditional methods,the proposed method establishes a higher resolution model, contains richer information,and has a high porosity in the distribution of sand bodies.Besides,the distribution of high values is broadly consistent with the distribution of well logging porosity.As a deterministic modeling method driven by seismic information,this method can perform reservoir characterization more accurately and has important significance for geological modeling.