Preface
This book is aimed at postgraduates, undergraduates and workers in industry who
require an introduction to geostatistics. It is based on seven years of courses
to undergraduates, M.Sc. students and short courses to industry, and reflects
the problems which have been encountered in presenting this material to mining
engineers and geologists over a wide age range, and with an equally wide range
of numerical ability. The book would provide the foundation of a course of
about 20 to 30 hours, or of a five-day short course.
The level of mathematical and statistical ability required is fairly
rudimentary; it is sufficient to be able to cope with concepts like mean,
variance, standard error of the mean, normal and log-normal distributions, and
to have some notion of the background to solving sets of simultaneous
equations.
As an introduction to a subject which is commonly presented as rather complex,
the book will familiarise the reader with the concepts and techniques of
geostatistics, providing the necessary foundation to enable him or her to
evaluate basic idealised examples. It also gives an indication of how to employ
the techniques in more complex and realistic situations.
Geostatistics is used throughout this book in its European sense of the ‘Theory
of Regionalised Variables', developed by Georges Matheron and co-workers at the
Centre du Morphologie Mathématique at
The units of measurement used in the book also reflect the state of the mining
industry. No attempt has been made to standardise to SI units.
The examples are real examples, and it seems absurd to turn, for example, 5-ft
cores into 1.52-m cores. The one case where this seems to
have been done is an actual example where the mine involved had adopted SI
units, but continued to use its pre-metric 5-ft core boxes.
In the presentation of the material I have tried to show how the basic ideas
may be developed intuitively, and I have tended to avoid supporting the ideas
with a rigorous mathematical derivation, since there are numerous existing
publications which use this latter approach almost exclusively. While many
computational difficulties can be eased by use of computer programs, such
assistance should not be needed within the scope of this text. None of the
examples is too unwieldy for pencil and paper, far less for a calculator. Where
formulae are too complex (or tedious) to calculate by hand, tables have been
provided. One example has been included (the simulated iron ore body) so that
some experience may be gained at tackling a fairly realistic example, to see
whether the reader can reproduce the author's result.
Acknowledgements are due to Richard Durham, who provided some of the examples
and the simulated iron ore body; Reg Puddy who produced the splendid drawings; Dr. C. G. Down
who created the situation which forced me to write the book; Malcolm Clark who
baby-sat and produced some of the prettier tables; and finally to André Journel and others at Fontainebleau who taught me all I
know about the theory of the Theory of Regionalised Variables. Any shortcomings
and inaccuracies in the text lie with me.
ISOBEL