Stratigraphic maps, such as structure, isopach, and facies maps, are commonly prepared as contour-type maps from measurement data obtained in wells and outcrops. In recent years geologists have experimented with methods for extracting additional information from these maps by application of various mathematical and statistical procedures. Most of the methods are very time-consuming and do not justify their cost as routine procedures without high-speed computers. The advent of computers has made possible a change in the entire framework of map preparation, analysis, and interpretation by furnishing quicker ways of assembling, storing, and processing the basic data. In this respect the computer and associated equipment act as a super-speed desk calculator and filing system hat frees the geologist from much busy-work and gives him more time to interpret and use his final maps.

Among problems that can be examined conveniently with the aid of computers are similarities or differences among maps; the use of maps as predicting devices; and the more general question of setting up criteria for the selection of mappable variables that will give the most information per dollar in terms of the objectives of the map study.

Map comparison and the use of maps as predicting devices can be achieved at reasonable cost by trend surface analysis, by which the "observed" map data can be separated into two main parts--the trend surface that represents the broad areal changes in the mapped variable, and the deviations from the trend that represent small-scale or anomalous variations. Sometimes the trend surface is of major importance in a study, but in some applications the deviations may rise to major importance. Selection of particular aspects for stratigraphic mapping can be approached in several ways--by regression procedures that "sort out" the important mappable variables; by multiple correlation procedures; or by use of factor analysis that identifies certain groups of variables as being of first-rank impo tance in the context of a map study.

Emphasis in this paper is on some principles that underlie map analysis, illustrated by sequential trend analysis of map data. The influence of open and closed number systems--rock thicknesses in contrast to percentages--is

End_Page 270------------------------------

developed by examples to show how map patterns are in part controlled by the ways in which the mapped data are expressed.

End_of_Article - Last_Page 271------------