Geomathematical and Petrophysical Studies in Sedimentology

The studies assembled in this volume cover a wide range of topics and applications. In search of communality in subject matter or method of analysis, five main central themes can be identified: computer-simulation models; quantitative methods for the analysis of petrophysical logs; seismic stratigraphy, and the application of trend-surface analysis and multivariate analysis to sedimentological and related problems. The articles have been arranged according to their relation to these central themes, in the mentioned order. The utilization of quantitative modeling techniques for decision making and the analysis of complex sedimentological phenomena is treated in five contributions, covering a diverse suite of applications. Arnon tackles the problem of computerizing the identification of terrigenous depositional environments. The proposed solution hinges on simulating the human logical evaluation process in which, based on empirically established associations between specific depositional settings and key sedimentary attributes, a systematic elimination procedure is employed to converge on the most plausible interpretation. The compilation and organization of the data on the association between sedimentary features and specific terrigenous depositional environments is an important contribution in its own right. Implicit to the entire approach is the multiple working hypothesis principle. The importance of colledting and reporting raw data in a systematic, objective, and consistent manner is duly emphasized. Arnons approach to the problem borders on certain methodologies developed by researchers in the field of artificial intelligence and it seems to offer various possible interesting extensions. In particular, it would be desirable to expand the system to encompass carbonate depositional environments. Also, the implementation of such algorithms as interactive systems could serve as an important learning and training aid. The spatial and temporal coincidence of all the geological factors which give rise to the deposition of a turbidite sequence is stochastic in nature. By assuming that bed thickness corresponds to the waiting-time variable in a stochastic Poisson process, Nishiwaki was able to obtain a reasonably acceptable agreement between bed-thickness distributions in simulated and observed turbidite sequences. However, as emphasized by the author, although the results are more satisfactory than those obtained previously by using a deterministic model, they are nevertheless deficient and unrealistic in several respects. This, no doubt, is due to the fact that in order to keep the model reasonably manageable from a conceptual and computational standpoint, certain oversimplifications are required. Furthermore, a number of important geological processes and parameters may be left out of consideration because we do not know as yet how to quantify them so as to be able to include them in the computerized-simulation model. This seems to remain an outstanding problem which reoccurs in many computer-simulation experiments in geology.