Evaluating the impact of data quantity, distribution and algorithm selection on the accuracy of 3D subsurface models using synthetic grid models of varying complexity

Testing the accuracy of 3D modelling algorithms used for geological applications is extremely difficult as model results cannot be easily validated. This paper presents a new approach to evaluate the effectiveness of common interpolation algorithms used in 3D subsurface modelling, utilizing four synthetic grids to represent subsurface environments of varying geological complexity. The four grids are modelled with Inverse Distance Weighting and Ordinary Kriging, using data extracted from the synthetic grids in different spatial distribution patterns (regular, random, clustered and sparse), and with different numbers of data points (100, 256, 676 and 1,600). Utilizing synthetic grids for this evaluation allows quantitative statistical assessment of the accuracy of both interpolation algorithms in a variety of sampling conditions. Data distribution proved to be an important factor; as in many geological situations, relatively small numbers of randomly distributed data points can generate more accurate 3D models than larger amounts of clustered data. This study provides insight for optimizing the quantity and distribution of data required to accurately and cost-effectively interpolate subsurface units of varying complexity.