Understanding the two-dimensional quantification of bioturbation intensity through computer modeling and statistical analysis

There is a common understanding among geoscientists that two-dimensional (2D) views of samples have their limitations in being able to represent the actual bioturbation intensity of sedimentary strata. However, a quantitative assessment of how 2D views might overestimate or underestimate bioturbation intensity (i.e., create errors in the estimation of bioturbation intensity) and a determination of what controls such errors is lacking. Exploring this knowledge gap with rock samples is challenging because it is not easy to find rock samples with a range of variables that control the reflection of bioturbation intensity in 2D views. To address this and to overcome the challenge of sample availability, this study generated 15 multipoint statistics models (MPS) of three burrow morphologies, including boxwork (Thalassinoides), vertical (Skolithos), and horizontal (Planolites) burrows. These MPS models allowed for the constraint of burrow attributes and bioturbation intensity and provided 9000 different 2D model slices to investigate how 2D views reflect actual bioturbation intensity. The error in the estimation of bioturbation intensity was calculated as the difference between the actual 3D burrow intensity of the MPS models and the bioturbation intensity estimated from the 2D slices extracted from these models. The results indicated that the chance of error in the estimation of bioturbation intensity when using a conventional nonlinear scale scheme was relatively low (< 14.4% and < 4.5% for bioturbation indices 2 and 3, respectively). Nonetheless, the results showed that, within each bioturbation index of these classification schemes, there was a substantial error in the estimate of bioturbation intensity if using a linear scale of burrow percentage (~ 10 to ~ 50, in 10% steps). A low percentage of the 2D views had a probability of reflecting the exact bioturbation intensity of the MPS models (~ 15%), and ~ 50% of the 2D views could almost retain the actual bioturbation intensity of these models, with values within the first standard deviation (4.5%). Approximately 35% of the 2D views overestimated or underestimated the actual bioturbation intensity by a value greater than the first standard deviation of the bioturbation intensity from the 2D views of the models. The results suggest that the error in the estimation of bioturbation intensity depends on the magnitude of the actual bioturbation intensity in the 3D MPS models and the orientation of the 2D slices of the models, but is independent of burrow morphology. The data and results presented here can be used to evaluate uncertainty regarding the quantification of bioturbation intensity from 2D views.