Reproduction assessment of connected geobodies in multiple-point simulation

Guillaume Rongier and Pauline Collon and Philippe Renard and Julien Straubhaar and Judith Sausse. ( 2013 )
in: Proc. 33rd Gocad Meeting, Nancy

Abstract

Multiple-point simulations (MPS) are booming stochastic simulation methods due to their ability to better take into account higher-order statistical structures. They are based on borrowing the multiple-point statistics not from the data but from an external representation of the expected geology. However, the reproduction of the geobodies from that representation is still a problematic topic. While a lot of methods try to improve the reproduction quality, few works have been done on the characterization of this quality. Currently, the quality control of a method is done using a visual assessment by comparing two or three realizations of the method with two or three realizations of some other methods. While the visual assessment lacks objectivity, it also complicates the check of an important number of realizations. If the reproduction issues are visually obvious with continuous geobodies such as channels, the existing indicators fail to give a precise information on the satisfactory aspect of a realization. In addition to classic indicators such as the facies proportions, we propose to check parameters linked to the connected geobodies of the realizations. Four categories of indicators have been defined to assess the reproduction of: i) The facies. ii) The connected geobody organization. iii) The connected geobody shape. iv) The connected geobody skeleton. These indicators allow to compare several images. To facilitate this comparison, we rely on the computation of dissimilarities between the images and their analysis using heat maps and two-dimensional representations based on multidimensional scaling. The methodology is then applied on several synthetic cases and associated realizations made with different simulation methods. Whereas multidimensional scaling is a powerful visualization tool, it induces some errors in the representation of the dissimilarities and should be only used for a first-order analysis. Details considering the relationship between the realizations and the methods should be preferably analyzed on the heat map as it represents directly the dissimilarities. If the visualization and analysis process of the dissimilarities is quite satisfying, further work should be done to improve the indicator capacity to capture the realizations characteristic.

Download / Links

BibTeX Reference

@inproceedings{RUNKJRM22,
 abstract = { Multiple-point simulations (MPS) are booming stochastic simulation methods due to their ability to better take into account higher-order statistical structures. They are based on borrowing the multiple-point statistics not from the data but from an external representation of the expected geology. However, the reproduction of the geobodies from that representation is still a problematic topic. While a lot of methods try to improve the reproduction quality, few works have been done on the characterization of this quality. Currently, the quality control of a method is done using a visual assessment by comparing two or three realizations of the method with two or three realizations of some other methods. While the visual assessment lacks objectivity, it also complicates the check of an important number of realizations. If the reproduction issues are visually obvious with continuous geobodies such as channels, the existing indicators fail to give a precise information on the satisfactory aspect of a realization. In addition to classic indicators such as the facies proportions, we propose to check parameters linked to the connected geobodies of the realizations. Four categories of indicators have been defined to assess the reproduction of: i) The facies. ii) The connected geobody organization. iii) The connected geobody shape. iv) The connected geobody skeleton. These indicators allow to compare several images. To facilitate this comparison, we rely on the computation of dissimilarities between the images and their analysis using heat maps and two-dimensional representations based on multidimensional scaling. The methodology is then applied on several synthetic cases and associated realizations made with different simulation methods. Whereas multidimensional scaling is a powerful visualization tool, it induces some errors in the representation of the dissimilarities and should be only used for a first-order analysis. Details considering the relationship between the realizations and the methods should be preferably analyzed on the heat map as it represents directly the dissimilarities. If the visualization and analysis process of the dissimilarities is quite satisfying, further work should be done to improve the indicator capacity to capture the realizations characteristic. },
 author = { Rongier, Guillaume AND Collon, Pauline AND Renard, Philippe AND Straubhaar, Julien AND Sausse, Judith },
 booktitle = { Proc. 33rd Gocad Meeting },
 location = { Nancy },
 month = { "sep" },
 title = { Reproduction assessment of connected geobodies in multiple-point simulation },
 year = { 2013 }
}