HR: 0800h
AN: GP21A-0103    [Abstracts]
TI: An improved algorithm for calculating first-order reversal curve (FORC) distributions using locally-weighted regression smoothing
AU: Harrison, R J
EM: rjh40@esc.cam.ac.uk
AF: Department of Earth Sciences, University of Cambridge Downing Street, Cambridge, U.K CB2 3EQ, United Kingdom
AU: * Feinberg, J M
EM: jfei05@esc.cam.ac.uk
AF: Department of Earth Sciences, University of Cambridge Downing Street, Cambridge, U.K CB2 3EQ, United Kingdom
AU: * Feinberg, J M
EM: jfei05@esc.cam.ac.uk
AF: Institute for Rock Magnetism, Department of Geology & Geophysics University of Minnesota, Minneapolis, MN 55455, United States
AB: First-order reversal curves (FORCs) are a powerful method for characterizing the magnetic hysteresis properties of natural and synthetic materials, and are rapidly becoming a standard tool in rock magnetic and paleomagnetic investigations. Here we describe a modification to existing algorithms for the calculation of FORC diagrams using locally-weighted regression smoothing (often referred to as loess smoothing). Like conventional algorithms, the FORC distribution is calculated by fitting a second degree polynomial to a region of FORC space defined by a smoothing factor, N. Our method differs from conventional algorithms in two ways. Firstly, rather than a square of side (2N+1) centered on the point of interest, the region of FORC space used for fitting is defined as a span of arbitrary shape encompassing the (2N+1)² data points closest to the point of interest. Secondly, data inside the span are given a weight that depends on their distance from the point being evaluated: data closer to the point being evaluated have higher weights and have a greater effect on the fit. Loess smoothing offers two advantages over current methods. Firstly, it allows the FORC distribution to be calculated using a constant smoothing factor all the way to the Hc = 0 axis. This eliminates possible distortions to the FORC distribution associated with reducing the smoothing factor close to the Hc = 0 axis, and does not require use of the extended FORC formalism and the reversible ridge, which swamps the low-coercivity signal. Secondly, it allows finer control over the degree of smoothing applied to the data, enabling automated selection of the optimum smoothing factor for a given FORC measurement, based on an analysis of the standard deviation of the fit residuals. The new algorithm forms the basis for FORCinel, a new suite of FORC analysis tools for Igor Pro (www.wavemetrics.com), freely available on request from the authors.
DE: 1540 Rock and mineral magnetism
DE: 1594 Instruments and techniques
SC: Geomagnetism and Paleomagnetism [GP]
MN: 2007 Fall Meeting