Abstract: With the increasing intensity of coal mining, the risk of mine water hazards has become more severe. As one of the key parameters of hydrogeological conditions in strata, porosity plays a vital role in mine water prevention and control. Accurately characterizing its spatial distribution is essential. Taking the Qinglongsi coal mine as a case study, a stratigraphic geometric model was constructed using discrete smooth interpolation (DSI) based on data from 140 boreholes, contour maps of the coal seam floor, coal seam thickness maps, and geological cross-sections. Among the boreholes, 66 contained logging data. For key stratigraphic layers, such as the main mining coal seam and its roof and floor, as well as major aquicludes, the model achieved a grid vertical resolution of 0.2 meters, with approximately 35 million hexahedral elements used to discretize the geological space. Based on this model, porosity curves were calculated by segmentally combining volumetric physical models with regression analysis using a density threshold of 2.3 g/cm³. These curves were then mapped onto the grid model. Constrained by a lithological model, the porosity attribute model was further generated using a sequential Gaussian simulation algorithm. The results show that by setting a density threshold, dynamically optimizing calculation parameters, and incorporating the lithological model as a geostatistical constraint, the porosity attributes within each lithological unit closely follow the distribution patterns of the original data. This significantly improves the accuracy of porosity interpretation and leads to the construction of a porosity model that aligns with lithological distribution characteristics. Furthermore, the porosity model of the working face roof can help predict variations in water-bearing capacity, providing data support for mine water hazard prevention. For instance, cross-validation between the porosity models at the 25 m and 40 m horizons of the 5-20303 working face roof and the transient electromagnetic detection results indirectly demonstrates the reliability of the model. Moreover, a significant correlation exists between high-porosity zones and low-resistivity anomalies, which can provide geological parameters and indicative information for evaluating aquifer enrichment in the roof. This study provides a valuable reference for the development of 3D geological property modeling and offers a new technical approach for analyzing hydrogeological characteristics in mines.