Reliability of the AARM Ellipsoid

Abstract:

Anisotropy of anhysteretic remanence (AARM) provides a unique tool to quantify geological fabrics. In contrast to anisotropy of magnetic susceptibility (AMS), it extracts information from the remanence carrying particles and is not prone to inverse fabrics. Compared to other remanence-based methods (like IRM or TRM), it does not alter the sample and allows one to focus on sub-populations by using partial ARMs. Additionally AARM experiments can be done on 99 samples per run with our fully automated SushiBar. However, a remanence is imparted through a sequence of reference directions and subsequently measured with a separate device, giving rise to potential alignment differences that mostly affect the off-axis components. Using vectorial measurements as delivered by common rock magnetometers is favorable to get the best estimate of the anisotropy tensor, but makes the result highly dependent on the exact knowledge of the reference directions. This can be overcome by using the projection of the remanence vector on the reference direction instead of the full vector in the calculation of the best-fit anisotropy tensor. Since AARM experiments are much more time consuming than AMS experiments, it is troublesome to compensate the loss in information with additional measurements. Therefore we propose how a projection method can be used to obtain an initial estimate of the original reference directions, which can then can be used to exploit the full vectorial calculations. We will show examples from simulations and measurements on rock specimens to demonstrate the effect and the improvements due to our new method.