Reassessment of some Holocene Sedimentary Paleomagnetic Records with Implications for Geomagnetic Field Models
Temporally continuous global spherical harmonic models of the Holocene geomagnetic field (e.g., CALS3k.4 and CALS10k.1b) rely on compilations of published sedimentary paleomagnetic records for their construction. In current models all data are initially included regardless of their quality and only extreme outliers are rejected during the fitting procedure. Encouragingly, they can extract globally and regionally consistent signals from the data; however, low quality paleomagnetic data and erroneous age models may distort geomagnetic field structures generated by the models. One particularly interesting non-dipolar feature observed in CALS3k.4 and CALS10k.1b is an undulation of the magnetic equator and associated paired flux patches at the core-mantle boundary under southeast Asia and northern Australasia. We re-examine and reconstruct a number of previously published records that influence this region. Although these records were suitably analyzed for the original aims of the specific studies, for global modeling it is desirable to treat data consistently wherever possible. Four problems commonly plague reconstructions: 1) the data used for modeling are often a smoothed composite from multiple cores, rather than horizon-level inclination, declination or relative intensity data from individual cores; 2) varied methods for correlation and smoothing lead to uneven data consistency; 3) advances in radiocarbon dating have resulted in changes to the calibration curve for atmospheric radiocarbon leading to possible offsets in time series across studies; and 4) age-depth models often do not fully consider the uncertainty distribution of the radiocarbon mixing profile and calibration process, producing implausible results. Reconstruction of a composite record is only possible when an author provides raw core data. This involves two key steps: re-correlation of data between cores and creation of a new age-depth model. In both cases we ultimately attempt a uniform approach. To aid correlation between core records from the same site we have constructed an algorithm that develops a conjoint signal to synchronize data sets independently from the geomagnetic signal recorded in the inclination, declination or relative paleointensity. We compare the results of this approach with the published correlations, our attempts at manual correlations and the outputs of some other commonly used signal matching programs. Age-depth models can be calculated in a variety of ways and we show differences in resulting time-series based on models ranging from linear interpolation between median calibrated radiocarbon ages to iterative Bayesian analyses of suites of calibrated radiocarbon ages.
AGU Fall Meeting Abstracts
- Pub Date:
- December 2012
- 1522 GEOMAGNETISM AND PALEOMAGNETISM / Paleomagnetic secular variation