Q'Gas: A Minicomputer-Based System to Aid in the Interpretation Of Exploration-Oriented Geochemical Data

Abstract

Lavin, O.P. and Nichol, I., 1981. Q'GAS: a minicomputer-based system to aid in the interpretation of exploration-oriented geochemical data. In: A.W. Rose and H. Gundlach (Editors), Geochemical Exploration 1980. J. Geochem. Explor., 15: 521–537. The past few years have seen an increasing application of computer-based procedures in the processing and interpretation of geochemical data. In many cases this has been carried out by non-geologists using large, general purpose computers remote from the exploration effort. In such situations, appreciation of the geological nature of the problems is often inadequate and in a number of cases misapplication of procedures has resulted. The present paper describes the Q'GAS system which consists of a minicomputer and a series of compatible interactive programs. These programs can be used independently by geologists who have a minimum of experience with computers allowing them the opportunity to carry out personally the processing and interpretation of geochemical data and to ensure that the methodology is based on a firm geological framework. The expense of such a system is relatively low: approximately $25,000 capital costs, and $3500 per year running costs including a maintenance contract. In addition approximately one man-month per year is required for supervision and up-keep. The system has the capability of: (a) selecting subgroups of samples that meet specified criteria (e.g. specific rock type or value ranges); (b) transforming data (addition, subtraction, multiplication, division, logarithms, exponentiation and random number generation); (c) making statistical computations; and (d) producing graphical displays (e.g. histograms, X-Y plots, symbol maps etc.). Flexible diskettes are used for data storage and communication with the system is achieved through a video terminal. Hard-copy output is produced on a small printer. The system presently includes programs for data management, univariate statistics with histograms, correlation analysis, X-Y plots, line printer symbol maps, line printer geochemical profiles, multiple linear regression, discriminant analysis, and R-mode factor analysis. Attention is presently being given to developing programs that use a pen plotter for producing better quality maps and diagrams at any scale. Experience has shown that the simpler programs for construction of maps, graphs, and diagrams can provide an immediate improvement in the quality, thoroughness, and speed of data interpretation as well as significantly reducing the tedium associated with manual methods. The multivariate statistical techniques, as always, require a higher level of expertise and many more man hours if they are to be used successfully.