1Kansas Geological Survey
2University of South Carolina
3Marathon Oil Co.
Originally published in 1991 as Kansas Geological Survey Bulletin 233. This is, in general, the original text as published. The information has not been updated.
This volume is the result of the conference "Sedimentary Modeling: Computer Simulations of Depositional Sequences," held at the Kansas Geological Survey in Lawrence, Kansas, on October 13, 1989. The idea to hold a conference was initiated in recognition of 100 years of continuous service by the survey and because of the recent advances in sedimentology that incorporate sequence-stratigraphic concepts, computer modeling, and related methods that help to make interpretations of the stratigraphic record more precise and accurate. Abstracts of the 32 papers presented at the conference were published as Subsurface Geology Series 12 by the Kansas Geological Survey. The 30 articles in this volume include papers from presentations at the conference and additional papers not presented at the conference but pertinent to the subject matter.
The goals of the symposium and of this volume were to include papers from a variety of geologic disciplines that focus on modeling and/or the collection and analysis of geologic data sets that have the potential to constrain quantitative sedimentary models. The idea was to facilitate further communication between groups with differing emphases and to heighten awareness on subjects that we believe will become increasingly important in sedimentologic studies. This volume should contain something of interest for field-oriented geologists and computer modelers alike, and we hope that it bridges the gap between these two disciplines.
Geologists, especially those involved in genetic stratigraphy, have long used their knowledge of geologic processes to predict lateral and vertical facies relationships in the field. The predictions were the result of qualitative "forward models" operating in their minds. By going to the field, they collected observations to test their models. Simulation of geologic phenomena has been a long-standing activity in geology. Analogue models, commonly scaled-down versions, approximate process-response relationships observed in nature (e.g., flume experiments for describing relationships between water flow and sedimentary structures, clay models for describing the mechanics of faulting and folding, and scale models of rivers and coastlines for examining perturbations of these systems). These types of simulations permit quantification, control, and repeatability for improved experimentation. However, the results of the analogue models are limited to simple systems. In contrast, computer systems do not have this inherent limitation and offer a broad, promising potential for simulations.
Computer simulation, dating back some 20 years in the earth sciences, has undergone rapid development in the physical and biological sciences in the last decade with the improvement of computer hardware and software, including the introduction of personal computers, the rapid increase in processing speed and memory, higher graphics resolution and broader color selection, more interactive and user-friendly systems, and increasing portability and adaptability of programs. The use of computer simulations in sedimentary geology provides the means to quantitatively evaluate the interaction and relative importance of numerous geologic processes and their constituent process variables. As Computers become more a part of the geologist's workplace, the quantitative assessment of alternative interpretations will be more common.
The earliest years in the rapidly evolving field of sedimentary computer modeling (represented in several modeling volumes) were dominated by algorithm development and software implementation. Workers in the modeling field spent most of their time building the models and much less time testing their models with field-based geologic data sets. This aspect may have served to create a rift between modelers and classical field geologists. We are now at the threshold of a new stage of modeling in which modelers view the collection of geologic data sets to test their models as an integral part of the discipline. The realization that models must be calibrated with ground-truth geologic data to gain credibility heralds a new stage in the evolution of computer modeling, a stage where field investigations and model simulations are viewed as complementary approaches to geologic problem solving.
Computer simulations are advancing toward more routine geologic applications and more sophisticated and complex interpretations. Although the building of computer simulations of sedimentary processes is instructive and valuable as a teaching tool, the real promise of computer modeling is in the area of prediction. Can we build sufficiently realistic computer models that will predict lithologic associations in areas with limited geologic control? For computers to gain acceptance, they must produce predictions that are not otherwise obvious or that are not easily quantified without computers.
Predictions derived from model simulations are only as good as the assumptions used in the model design and model input. The range of uncertainty in these assumptions is a measure of the uncertainty of the result. This is similarly true of predictions made with traditional qualitative model projections. In areas of limited geologic control, a successful comparison of model output to geologic data points simply shows that the combination of model input parameters used is consistent with the known geology of the study area. However, the model-data match does not prove that the used combination of modeling parameters is uniquely correct. Consequently, predictions of lithologic associations in areas with little geologic control will have an associated uncertainty related to the uncertainty in defining the ranges of model input variables. Several hypotheses for the geologic history of an area should be considered to explain the known distribution of geologic data points, because each hypothesis will have different implications in any prediction scheme.
An advantage of computer simulations over conventional qualitative methods is that the uncertainty in any prediction can be quantified if the uncertainty in model input can be statistically defined. Several approaches are described in this volume. The theoretical range of any particular geologic parameter (e.g., subsidence rates) may be broad, but the geologist should be able to define the most probable distribution for that parameter. By defining and constraining parameter-range distributions, geologists can put boundaries on the possible number of reasonable solutions in a given model experiment. Geology has always suffered as a science because of our inability to define the uncertainty of our conclusions. The ability to address quantitatively the uncertainty in a geologic interpretation or prediction signals a new age for geology as a discipline.
Sedimentary computer models reflect a simplified mathematical representation of geologic processes operating on geologic time scales. At any given point in time they represent our current thinking of how geologic systems work. Models in geology evolve when predictions from these models are compared to observations from the geologic record. Likewise, computer models must evolve through an iterative process whereby model predictions help to focus new field observations, which in turn provide constraints for improved model designs.
Ultimately, computer modeling of sedimentary processes should be viewed as a tool for stratigraphers and sedimentologists to constrain geologic interpretations and predictions. This tool allows for the quantification of both theoretical and empirical constraints on geologic processes and ensures that interpretations and predictions are consistent with these constraints. To take advantage of the power of computer modeling, geologists must train themselves to become aware of the types of observations that can be used to constrain model predictions. Conversely, modelers must continuously reconcile their models with real geologic observations.
In the future, quantitative evaluations of geologic interpretations and predictions using computer models may become commonplace. Communication between groups of differing expertise (e.g., sedimentologists, petrophysicists, petrologists, paleontologists, and modelers) will become increasingly important for sharing perspectives on how to improve observational frameworks and strategies for simulating geologic processes. This volume is dedicated to the spirit of that communication. We encourage future workers to follow a similar path.
Special thanks go to M. Braverman for editing and arranging manuscripts into their final format and to J. Sims for drafting and fine-tuning the figures and for designing the cover. We appreciate the efforts of M. Adkins-Heljeson and R. Buchanan in expediting the publication of this volume. We are indebted to the people listed below for their time and expertise in reviewing the papers included in this volume. Finally, we are grateful to the authors for their efforts in the preparation and completion of the manuscripts.
|A. W. Archer||C. Frohlich||R. K. Matthews||L. L. Sloss|
|D. L. Baars||R. K. Goldhammer||J. May||D. G. Smith|
|W. M. Benzel||S. M. Greenlee||T. Mazza||S. Snelson|
|A. L. Berger||J. Grotzinger||S. J. Mazzullo||S. Speyer|
|D. Bice||M. T. Harris||D. McCubbin||M. Steckler|
|D. Boardman||P. M. Harris||W. J. Meyers||G. Stevenson|
|G. Bond||W. Harrison||D. Osleger||G. Stockmal|
|R. Brenner||L. Hinnov||C. Paola||S. Stonecipher|
|C. Brett||J. Imbrie||W. Pitman||D. M. Tetzalaff|
|B. J. Bunker||R. Inden||H. Posamentier||J. A. Thome|
|R. R. Charpentier||R. L. Kaesler||B. Rascoe||H. Wanless|
|P. Choquette||G. D. Klein||J. F. Read||J. Warme|
|J. Davis||M. Kozar||P. Santogrossi||D. E. Watts|
|R. Demicco||E. Kvale||C. Savrda||R. Weimer|
|S. L. Dorobek||M. Lagoe||J. W. Schmoker||R. R. West|
|P. Enos||D. Lawrence||S. Schutter||B. Witske|
|R. Farrar||M. Longman||K. Shanley||P. Wong|
|A. Fischer||C. G. Maples||A. Simo||P. Yarka|
Kansas Geological Survey
Comments to email@example.com
Web version updated April 30, 2010. Original publication date 1991.