Mathematical Geology, Vol. 7, No. 4, 1975
Reviews Computer-Based Storage, Retrieval, and Processing of Geological Information edited by R. Bergeron, S. C. Robinson, and C. F. Burk, Jr. International Geological Congress, 24th Session, Montreal, Section 16, 1972, 222 p. This is a collection of 26 papers devoted to computer processing of geological information. In fact, it marks the first session in any International Geological Congress devoted exclusively to the use of computers in information processing. The 26 papers are organized into six subsections, which are devoted, respectively, to (1) geological field data systems, (2) geophysical and geochemical data systems, (3) paleontological and oil-well data systems, (4) data systems as applied to ore deposits, mineralogy, and petrography, (5) evolution, evaluation, and general application of machine data files, and (6) bibliographic application of machine data systems in the geosciences. Collectively, these 26 papers provide an overview of the state-of-the-art in computer methods of storing, retrieving, and manipulating geological data. Because the papers describe procedures and applications from a wide diversity of sources, however, it is difficult to comment on the collection as a whole without first surveying the papers individually. The initial paper in the first group (by Berner, EkstrSm, Lilljequist, Stephansson, and WikstrSm) deals with the GEOMAP system, which is a system developed in Sweden for plotting geological data on maps, as for example, plotting of strikes and dips, or rock types at specific point localities on the map. The system can be used to plot virtually any type of data on a point-by-point basis. Contrary to its name "GEOMAP," however, the system is not applicable for plotting conventional geologic maps in which different map units are shown on the map. The system seemingly has proven to be useful in Scandinavia. Another paper (Laurin, Sharma, Wynne-Edwards, and Franconia) in this same subsection deals with a system developed by the Quebec Department of Natural Resources for machine handling of data collected at outcrops. The system provides for systematized recording of information in the field that are to be keypunched subsequently. Data thus 355 9 1975 Plenum Publishing Corporation, 227 West 17th Street, N e w York, N . Y . I0011. N o part o f this publi .c~tion m a y be reproduced, stored in a retrieval system, or trans .r.nitted., in any f o r m or by.any means, electronte, mechanical, photocopying, microfilming, rzcording, or otherwise, w~thout wrttten permlss~on o l the puonsner.
356
Renews
transcribed can be automatically plotted on maps. The system involves codes for a variety of geologic features observed at outcrops, including outcrop size (an important factor in much of Quebec), rock type, texture, color, presence and size of folds, types of lineation present, presence of glacial striae, and so on. The system has been developed with close attention to its ultimate use in point plotting of data. For example, the field recording forms provide for clarifying the rock type at each outcrop into one of a standard set of lithologies that eventually will be used as geologic map units. The computerized system allows for considerable variation in rock texture, mineralogy, color, and grain size, but it may override the rock name applied by the field geologist if the name he applied is not in accord with the appropriate rock type as prescribed in terms of its mineral composition and other properties. A third paper (Roddick and Hutchison) in this subsection deals with a system for handling geological field data in the coastal regions of British Columbia, and is similar to the Swedish and Quebec systems. The remaining paper in the first subsection (Colonell, Goldsmith, and Turbide) deals with an automated plotting of beach profiles and wave-refraction diagrams. Two of the papers in the second subsection deal principally with plotting of chemical data on graphs. One deals with water analysis data (Dall'Aglio and Gigli) and the other (in French, by Pelet and de Jekhowsky) with quantitative rock-analysis data. The other two papers in this subsection deal briefly with handling of geophysical data obtained on shipboard (Shih and Heftier), and with digital processing of airborne infrared image signals (Stingelin and Avis). Two of the papers in the third subsection deal with paleontological applications. Berry describes the system in use at the University of California's Museum of Paleontology at Berkeley. Subsequently, Kent described the Colorado School of Mines' computerized data bank for foraminiferal data in the Western Interior of North America. The two other papers in this subsection deal with Saskatchewan's computerized well-information system (Buller) and with the use of a data system to statistically analyze the geometrical properties of Alberta oil and gas pools (Dickie and Williams). Papers in the fourth subsection deal with automated cartography involving the production of multicolored geologic maps (Bickmore and Kelk), calculation of map coordinates (Grandclaude, in French), treatment of metallogenetic province data (Lafitte, Capitant, Lenci, Teil, and Vannier, in French), and treatment of petrologic data (Harrison, Lawson, Hawkes, and Dangerfield). The fifth subsection includes papers dealing with computerized storage and retrieval of geologic documents (Haworth and Sparkes), a description of the University of Oklahoma's Generalized Information Processing System
Reviews
357
(GIPSY), by Moody, a brief note about a retrieval system in use in Russia, a generalized survey of the industrial use of computerized data files in oil exploration (Robinson), and a review of the evolution of oil-exploration data systems both in terms of their technology and adaptation to user's needs (Stauft). The final subsection contains papers dealing with computerized bibliographic systems, including those in use in France (Delbos and Dumont, in French), the Canadian System for Geoscience Data (Sharp), Geo-Archive, a commercial geoscience system in Britain (Lea), and the Smithsonian Institute's Science Information Exchange (Riva). In spite of the variety of topics treated in this report, most of them are treated superficially. The authors have not had the space to do an adequate job of describing their particular systems. This is unfortunate, because the interested reader has almost no other place to turn to really get an in depth look at data storage and retrieval systems as applied to the earth sciences. In short, then, this report is a useful point of departure, but few readers will find satisfaction if they seek a detailed description of systems potentially adaptable to their own needs. John W. Harbaugh
Department of Geology Stanford University Stanford, California (USA) Biometry by R. R. Sokal and F. J. Rohlf W. H. Freeman and Company, San Francisco, 1969, 776 p., $14.00 (US) As the title implies, this book describes the use of statistics in the life sciences. Apart from its biological context, it is basically an elementary statistics textbook, and in their Preface the authors are apologetic for adding one of their own to the deluge of elementary statistics books on the market. However, as the authors later demonstrate, the book has special features not customarily found in introductory statistics texts. The special features include strong orientation toward practical applications, use of real, copious rather than short, fictitious data in sample problems, a set of F O R T R A N programs designed to help implement basic statistical techniques, a tabular guide that shows which statistical methods are suitable for which type of problem, and a description of the use of desk calculators. Solutions to the sample problems are illustrated in special tables called "boxes" where the computations are carded out step by step. Statistical
358
Reviews
theory is kept to a minimum, and the mathematical background expected of the reader is that of college algebra. Most of the topics are fairly standard for an elementary text: descriptive statistics, probability distributions (mainly binomial, Poisson, and normal), estimation and hypothesis testing, regression, correlation, analysis of variance, and analysis of frequencies. However, analysis of variance techniques receive special attention; they also are introduced at a relatively early stage in the book. In addition, there is a discussion on the types and handling of data in biology and a mathematical appendix that shows the derivation of important formulas. Nonparametric tests receive more than a perfunctory coverage, as does the subject of data transformation. Certain advanced topics are mentioned briefly, and the reader is referred to appropriate books and papers for details and applications. Altogether, the book contains 17 chapters and 4 appendices. No statistical tables are provided in the book. Instead, the authors, under the conviction that placement of statistical tables in a textbook (even as an appendix) causes inconveniences to the reader, have chosen to furnish such tables under a separate cover (Statistical Tables, by F. J. Rohlf and R. R. Sokal, W. H. Freeman & Co., 1969, $7.50 text edition and $2.75 paper edition). Although an argument to the contrary can be made concerning inconvenience, the large volume of material included in the Tables seems sufficient reason to keep it separated from the main text. In any situation, the Tables should be considered a supplement to the main text. The book has few flaws. Employment of real, copious data in sample problems, although it has advantages, carries with it the risk of discouraging the less-than-enthusiastic student on account of the laborious computations involved. The mathematical appendix serves a good purpose, but the material in it could have been more profitably integrated into the text without materially affecting the practical bent of the book. Also, basic probability theory deserves more attention than it receives in the book. Finally, an introductory chapter on "modern" Bayesian inference would have given a refreshing alternative to the more classical approach based on sampling theory. Despite these weaknesses, the book has much to its credit. It is well written, well organized, well indexed, and richly "illustrated" with examples. The book has a high didactic value, both for self study and for teaching purposes. Despite its emphasis on applications, it does not have the superficiality of a do-it-yourself manual. The techniques introduced are discussed at length, mostly in nonmathematical terms. The material covered is more than one customarily finds in an elementary statistics book, and a few special topics covered are especially welcome. The F O R T R A N programs provided in the book are valuable in themselves, and can be used to build a modest statistical library. With all these features, Biometry would be well
Renews
359
suited as a textbook and as a reference book. Its greatest appeal should be to those students and researchers who are interested in applied statistics and who have a limited background in mathematics. Although the intended audience of the book is biologists, geologists would almost equally benefit from it. At the price of $14.00, Biometry could be considered a bargain. F. Demirmen
Shell International The Hague (Netherlands) Directory of Computerized Data Files & Related Software
(available from federal agencies 1974) NTIS (National Technical Information Service) NTIS-SR-74-01 U.S. Department of Commerce, 1974, 166 p., $60 (US) A directory of computerized data files is available now from NTIS (5285 Port Royal Road, Springfield, Virginia 22151). The directory is a guide to "machine-readable data files, data bases and related software available to the public from 60 Federal agencies." Included in the directory is information on about 500 data files. Each entry contains the title, date, generating agency, availability, and a brief abstract. Entries are grouped into 72 subject fields under three main headings: (1) Demography, Social Sciences, and Government, (2) Economics, and (3) Science and Technology. The subject field of special interest to geologists is Earth Sciences. Other categories of potential interest include: Environment and Geography; Atmospheric Sciences; Computers, Control and Information Theory; Geocoding and Dictionary Files; Libraries and Information Science; Material Sciences; Mathematical Sciences; Ocean Science and Technology; and Simulation and Models. Examples of entries in the Earth Sciences subject field are: Oil Shale Project, 1969 to present Watershed Conservation and Development Master File--1973, updated monthly Analytic Modeling of Rock Structure Interaction--Computer Program Magnetic Tape Containing Stream-Sediment Analytical Data, Central Maine Magnetic Tape Containing 7,192 Chemical Analyses of Sedimentary Rocks of the Central and Northwestern States Analyses of Natural Gases, Tape for 1917-1972 Cartographic Automatic Mapping Program (CAM), Final Rept. World Data Bank I In addition to the 72 subject fields, the entries are cross indexed by generating agency, subject, and number. The generating agencies for entries listed in the
360
Reviews
Earth Sciences subject field are Bureau of Land Management, National Geophysical and Solar-Terrestrial Data Center, National Weather Service, Department of Defense, Geological Survey, Bureau of Mines, and Department of Interior. In view of the large amount of material known to be available, it is surprising that so few entries are included. For example, none of the Geological Survey Computer Contributions or their computer-processable files such as CRIB and RASS are listed. The directory was the first largescale undertaking of this type, however, and maybe the next edition will have fuller coverage and be more complete. The items described in the directory can be ordered directly from NTIS or from the generating agency. NTIS provides a follow-up service to directory customers; for example, they will give or help locate additional information as needed. All data in the files meet the confidentiality requirements of the g~nerating agency. It is expected that the directory will be revised annually and eventually will be expanded to include all Federal agencies; the next edition is expected in 1975. This first effort is a notable one but hopefully the second edition will be even better, enhancing its value to geologists worldwide. D. F. Merriam
Department of Geology Syracuse University Syracuse, New York (USA)