ABSTRACT Threedimensional geologic parameter modeling is crucial foundational work in the fields of subsurface resource exploration and environmental protection. However, the traditional interpolation method has obvious limitations in dealing with the spatial variation of complex geologic bodies and sparse sample data. This work suggests a 3D modeling approach that combines Markov chain modeling and geostatistical inverse distance weighted (GIDW) interpolation to increase the vertical continuity of geological parameters and the accuracy of spatial predictions. To achieve the dynamic adaptive adjustment of the inverse distance‐weighted interpolation p th power and to finish the two‐dimensional water content planar interpolation, the neighborhood index is first computed using the borehole data. Three types of membership functions—triangular, trapezoidal, and Gaussian—are then constructed on the basis of the spatial distribution among sample points. Then, using the two‐dimensional interpolation findings as the initial state, the Markov chain model is utilized to perform the transfer probability statistics and state discretization, and layer‐by‐layer derivation is used to construct the three‐dimensional geological parameter model. According to the experimental results, the GIDW method's root mean square error (RMSE) is lower than that of the traditional inverse distance weighting (IDW) and Kriging methods by 12.4% and 18.7%, respectively, in the 2D interpolation stage; in the 3D modeling stage, the Markov chain successfully preserves the state continuity between the layers, and the profile prediction's RMSE is roughly 2.5. The current approach offers a good model for complicated subterranean environments and demonstrates strong accuracy and adaptability when handling sparse samples, vertical state progression, and geologic body heterogeneity. The approach offers efficient technological assistance for 3D modeling and resource prediction in intricate subterranean environments, and it demonstrates good adaptability and accuracy when handling a variety of geological bodies, sparse samples, and vertical state changes.
Bao et al. (Sun,) studied this question.