Automated biostratigraphic correlation of palynological records on the basis of shapes of pollen curves and evaluation of next-best solutions

Uncertainty management in stratigraphic well correlation and stratigraphic architectures: A training-based method

“…Further work includes making sensitivity analysis of the different parameters, studying the impact of the correlation order of the wells, and applying the method on real data. The DTW has been modified by several authors to implement a stochastic version of the algorithm (Nakagawa and Nakanishi, 1988;Pels et al, 1996). A possibility to reflect uncertainties of the cost computation is to randomly sample the propagation in the cost matrix (Lallier et al, 2009).…”

“…This paper builds on automatic stratigraphic well correlation methods, which make up an alternative option to generate consistent well correlations and to assess the associated uncertainties (Agterberg, 1990;Pels et al, 1996;Lallier et al, 2012;Agterberg et al, 2013;Lallier et al, 2016). Automatic correlation is reproducible, honors spatial observations and mitigates the subjectivity of interpretation.…”

“…In the case of regularly sampled well logs, the correlation coefficient within a depth window (Olea, 1994(Olea, , 2004 or the Lp distance norm between log values (Gordon and Reyment, 1979;Wheeler, 2015) may be used for this purpose. As these costs do not fully reflect the reasoning made by expert-based correlation and do not easily account for sedimentary gaps, several authors propose to use a zonation of wells in terms of intervals and a specific distance function that reflects differences between interval features such as lithology and thickness (Smith and Waterman, 1980;Howell, 1983;Waterman and Raymond, 1987;Fang et al, 1992;Pels et al, 1996). Recently, Lallier et al (2016) extended these rules for correlating carbonate platform and ramp sediments using stratigraphic sequences and paleo-bathymetry interpreted along the wells.…”

Computer-assisted stochastic multi-well correlation: Sedimentary facies versus well distality

“…However, the very large number of possible scenarios makes their manual validation almost impossible. To overcome this limitation, a numerical method to provide automatically multiple stratigraphic scenarios is proposed (e.g., Pels et al, 1996;Lallier et al, 2016;Edwards et al, 2018). The scenarios are consistent with the input data and with some explicitly defined stratigraphic concepts, which guarantee that the solutions are acceptable.…”

“…To retrieve such isochronous surfaces from stratigraphic series, we rely on the Dynamic Time Warping algorithm (Section 2). This numerical tool computes the likelihood of a correlation set among all possible correlation sets and returns the best correlation set, i.e., the most likely correlation set (Sakoe and Chiba, 1978;Waterman and Raymond, 1987;Hale, 2013;Wheeler and Hale, 2014) or the n-best correlation sets (Pels et al, 1996;Lallier et al, 2016;Edwards et al, 2018). The likelihood of one correlation set is computed by minimizing the cumulative cost of correlations and transition between correlations.…”