A NEW METHOD TO DETERMINE PALEOSECULAR VARIATION

This study is based on synthetic virtual geomagnetic pole distribution s simulated by the sum of a Fisherian distribution characteristic of p aleosecular variation (PSV) and a uniform distribution representing tr ansitional data resulting from reversals and excursions of the geomagn etic field. Using such simulations, the optimum cutoff angle to apply to estimate the most precise value of the angular standard deviation ( ASD) characteristic of PSV is determined for variable ASD and random d ata percentage. A relation between this optimum cutoff angle (A) and t he expected ASD is deduced: A(degrees) = 1.8ASD(degrees) + 5 degrees. A recursive method for determining the characteristic ASD using this r elation is proposed. It is compared with the classical selection metho d using a constant cutoff angle equal to 40 degrees. The method propos ed in the present paper gives much more regular results and especially corrects bias on low and high ASD values obtained with the classical method. The proposed method should be important for PSV modeling becau se the lowest and highest ASD are the most constraining. The relation cited above could also be used in magnetostratigraphy to limit the dom ain of transitional geomagnetic poles.