Evaluating Interpolation Methods by Geostatistical Modeling of the Douala Oil Field Porosity Data (Cameroon)
The choice of an interpolation method to re-sample and solve the problem of scattered data is often difficult, as several methods show large differences in results. In this study, we re-sampled the sparsely porous data using three interpolation techniques: the inverse distance to a power, minimum curvature and kriging. The experimental variogram of field data was generated. The anisotropy of these data has been simulated by the Gaussian model and we have concluded that these data present a geometric anisotropy. The interpolated variograms from the three techniques were plotted and fitted by the least squares method. These variograms gave a better accuracy and consistency than the field data variogram. The accuracy and performance of the interpolation techniques were evaluated by calculating their Variance, their Skewness, their Kurtosis and their Root-mean-squared. Contour maps and Wireframes were also computed from interpolated grids to perform a visual analysis of the porosity spatial distribution. Porosity distribution of the studied field has a higher continuity in NE-SW direction.