INTERSECTIONS OF REMANENCE SMALL CIRCLES: NEW TOOLS TO IMPROVE DATA PROCESSING AND INTERPRETATION IN PALAEOMAGNETISM

Abstract

Intersections of remanence small circles are increasingly used in palaeomagnetism to determine the palaeofield direction from synfolding remanences. The approach ['small circle intersection (SCI) method'] presupposes that remanence small circles from sites tilted to different directions intersect in a common point, which is assumed to represent the palaeofield direction. The SCI method is not restricted to synfolding remanences. It can be applied also to derive the palaeofield direction from pre-folding remanences, in addition to tilt correction and as a cross-check for fold tests. However, the SCI method has specific requirements and does not work in all cases. The suitability of data sets and reliability of palaeofield estimates have to be assessed carefully and always against the background of local geology. This paper examines the occurrence and distribution of SCI as a tool of judgement: first, a data set is most suitable and a palaeofield estimate by SCI most reliable, when most of the theoretically possible intersections are realized and, in addition, clearly grouping in a narrow region. Second, the small circles that greatly determine the result of SCI, are identified by the highest number of intersections they have with the others. These criteria allow a systematic investigation of SCI estimates upon stability and dependencies. Moreover, SCIs can be an indicator of irregular, unexpected tectonic settings. When the intersections disperse strongly although significant bedding strike differences are given, relative rotations or other unexpected displacements are likely to have occurred. If so, the SCI estimate will be invalid and probably biased, however, also Fisher mean calculation and most fold tests will be affected. In this case, the SCIs can help to separate the data into subsets valid for Fisher statistics and SCI methods. The other possibility is to use tentative reconstructions to infer the true sequence of tilts and rotations and correct the data adequately. © 2006 The Authors Journal compilation © 2006 RAS.