Improved Identification of Clastic Depositional Environments by Process and Facies Models in Expert System: ABSTRACT

Abstract

A previously described expert system for identification of clastic depositional environments (XEOD) has been substantially redesigned and improved as XEOD-II. Whereas the original design utilized simple checklist relationships between environments and features (sedimentary structures, lithology, etc), the new version is more meaningful in that it uses features to characterize facies as intermediate steps to an environmental interpretation. Some facies correspond directly to particular environments (e.g., tractive unidirectional currents). Facies classification is hierarchical, and identification proceeds to the most specific level allowable by the input data. More contexturally accurate interpretations for coastal and fluvial systems are possible using a series of data sets representing facies in a vertical sequence. Tentative facies identifications are verified by imperfect matching against a set of idealized vertical-sequence models. Incomplete and imperfect matches are diagnosed using alternative interpretations which combine two or more models. In spite of considerable difficulty in accepting both specific and varietal terms, improvements in the user interface have resulted in a relatively forgiving program. XEOD-II thus represents a second-generation system utilizing both factual knowledge and logical reasoning in sedimentological facies analysis.