Deconvolution Method for Directions of the Post-Depositional Detrital Remanent Magnetization Using an Exponential Fixing Function.

Abstract

A simple deconvolution method of the post-depositional DRM (detrital remanent magnetization) is proposed and examined using the synthetic data. Assuming the exponential fixing function of magnetic moments in sediments, the original field is expressed by a simple function of the convolved field. This leads us to the deconvolution of the post-depositional DRM if the fixing depth is known. In general the fixing depth is not given, so that the problem becomes complicated. However, adopting a few reasonable assumptions, the deconvolution is performed as a nonlinear problem with the fixing depth being an unknown parameter. Consequently the original field including the relative field intensity can be estimated from the direction data sets with the new method.

Further application of the deconvolution method of post-depositional DRM to the precise record of the Matuyama-Brunhes reversal in the sediments from the Boso Peninsula, Japan

“…The shortperiod variations of about 100 years are clearly recognized in both of the inclination and declination after the deconvolution. As shown by Tsunakawa (1995), the frequency response of the present deconvolution is very smooth and then any specific frequency component is not newly generated. Therefore these variations are believed not to be an artifact.…”

“…As the pDRM origin is inferred for the Boso sedimentary magnetism from the evidence of intensive bioturbation (Okada and Niitsuma, 1989;Tsunakawa et al, 1995), we have to restore more accurate variations from the measured remanence directions. Tsunakawa et al (1996) applied the deconvolution method (Tsunakawa, 1995) to the two directional datasets of about 900 years from the different sites, Yanagawa and Heizogawa. As a result, the half fixing depth of an exponential fixing function in pDRM acquisition (Hyodo, 1984) was estimated to be 21 ± 1 (1σ ) cm for the Boso sediments, which seems reasonable in comparison with those for other marine sediments (Tsunakawa et al, 1996).…”

Deconvolution of Paleomagnetic Directions from Marine Sediments in the Boso Peninsula, Japan, and Its Implications for the Matuyama-Brunhes Transition Field.

“…The Bayesian smoothing method for the paleomagnetic directions (Tsunakawa, 1992) is suitable for the Boso data sets because an abrupt change in the declination occurs during the reversal. For the Yanagawa data, smooth curves have already been obtained by the Bayesian procedure and are used in the present analysis (Fig.…”

“…First, we smooth the remanence directions from two sites of the Boso marine sediments by the Bayesian method (Tsunakawa, 1992). Second a new deconvolution method is applied to the smoothed directions so as to obtain three components of the geomagnetic field during the Matuyama-Brunhes reversal.…”

“…This error may be attributed to the remaining noise in the data sets. As pointed out by Tsunakawa (1992), it is difficult to completely remove noise by smoothing, especially, near the end portion of the sequence, and any remaining noise is amplified after deconvolution . Another cause may be the short period variations in the geomagnetic field, which cannot be detected at Heizogawa but are recorded at Yanagawa as mentioned earlier.…”