| VGP and VDM data are widely used in paleomagnetism to describe paleofield behavior, during times of stable magnetic polarity as well as during reversal transitions. Both of these concepts neglect non-dipole fields and, thus, the presence of non-dipole fields will lead to uncertainties. We introduce a new method, based on geomagnetic intensities, to calculate what percentage of the field is due to non-dipole fields, and find a value of 20–25% for the present and 17–18% a century ago. A model, based on the present field, is then used to investigate how VGP and VDM are affected by a relative increase of the non-dipole field. For a decreasing dipole, the VGP scatter increases while, surprisingly, the VDM scatter generally does not. The spatial distribution of VGP is used to simulate a magnetic reversal transition. There is large spatial variation in transitional VGP suggesting that reversal transitions and their duration, if defined by VGP latitude, depend on location; furthermore, the Brunhes reverse VGP are unlikely to have occurred from a simple reduction of the dipole moment with all other factors kept invariant.; An existing paleointensity database is expanded by about 30% and used to explore questions regarding the long-term paleointensity and paleosecular variation during the last 320 Myr. We introduce a method to use VDM scatter as a proxy for long-term paleosecular variation from intensities, which might help to constrain geodynamo modeling. An analysis of the VDM data suggests a correlation between the mean intensity and the VDM scatter which could be due either to a bias in the data (e.g. a rock magnetic artifact) or reflect true paleofield behavior. We find evidence for bimodality in the evolution of the VDM histograms. Surprisingly, the two peaks are almost constant with time, at around 4–5 × 1022Am2 and a value close to the present field. The bimodality might be the result of paleofield behavior, suggesting the presence of local energy minimum (LEM) states in the geodynamo, or could be due to a bias in the data. At present we cannot distinguish between these two possibilities. |
