Premium
STATISTICAL MULTIVARIATE ANALYSIS OF AIRBORNE GEOPHYSICAL DATA ON THE SE BORDER OF THE CENTRAL LAPLAND GREENSTONE COMPLEX *
Author(s) -
LANNE E.
Publication year - 1986
Publication title -
geophysical prospecting
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.735
H-Index - 79
eISSN - 1365-2478
pISSN - 0016-8025
DOI - 10.1111/j.1365-2478.1986.tb00516.x
Subject(s) - principal component analysis , geology , geophysics , linear discriminant analysis , magnetic anomaly , cluster (spacecraft) , multivariate statistics , statistical analysis , mineralogy , statistics , mathematics , computer science , programming language
Statistical multivariate methods for the integrated processing of airborne geophysical data were tested. The data consisted of magnetic, electromagnetic and gamma radiation measurements, to which cluster analysis, principal components analysis and discriminant analysis were applied. Also, auxiliary variables were derived from the original ones and their value was tested. Although the frequency distributions of the data do not favour statistical analysis, the practical results are acceptable. Principal component analyses show geological and technical aspects that are difficult to obtain from the original observations. In cluster analyses, the sources of measured fields control the grouping of variables. Discriminant analysis was applied to the automatic identification of rocks by geophysical data. The rocks investigated are metasediments and metavolcanics, some magnetic and others conductive. When all available geophysical data were included, correct identifications were made in more than 60% of cases. In particular, gamma ray observations were found to improve the discrimination of non‐magnetic and non‐conductive rocks. The geophysical similarity of rocks studied by cluster analysis depends on electrical and magnetic properties as well as on their origin; the content of radioactive elements in turn is related to the origin.