| In this study, a recently developed stochastic model of single domain (SD) paleointensity
behaviour is expanded to investigate the effects that anisotropic and non-linear
thermoremanent magnetizations (TRMs) have on the paleointensity results and the
parameters used to select data. The model results indicate that before applying any form of
correction these non-ideal factors can produce results that are self-consistent, but highly
inaccurate. The methods that are currently used to correct for anisotropic and non-linear
TRMs are effective and greatly increase the likelihood of obtaining accurate results. The
corrections, however, do not restore the results to those of ideal SD samples measured with
the same laboratory-to-ancient field ratio, but the data are restored to those of ideal SD
samples with the equivalent laboratory-to-ancient magnetization ratios (MLab-MAnc). The
simulations indicate that non-linear and anisotropic TRM have no or only a weak influence
on the parameters commonly used to select paleointensity data, which means that these
non-ideal factors are effectively undetectable. Given the high self-consistency and highly
inaccurate results that anisotropic and non-linear TRM can yield, it is essential to
test for such effects and all Thellier-type paleointensity studies must include tests
for anisotropic and non-linear TRM to assert the reliability of the data obtained. |