A hybrid approach to integrate multi-source geophysical data for inter-well formation property estimations

For petroleum exploration and development, inter-well formation property estimation is very important since it is the foundation of further reservoir modeling and simulation. For most cases, this task is performed based on property observations at well-points, while seismic data is also provided as the supplement. In essence, the inter-well formation property estimation is a spatial estimation task based on multi-source data. Even though various geo-statistical interpolation and machine learning mapping algorithms have been proposed, they all have limitations in estimation accuracy, horizontal resolution or algorithm assumption. In this article, through the combination of machine learning mapping and geostatistical interpolation, we propose a novel approach for better inter-well formation property estimation. The proposed approach is applied to a real-world inter-well shale volume estimation task for demonstration. Compared with existing methods such as ordinary kriging interpolation, co-kriging interpolation or machine learning mapping, the proposed approach shows significant advantages in estimation accuracy and hori-zontal resolution, which indicates that the proposed approach provides an alternative way for further inter-well formation property estimation practices.