Abstract
Recovering the geomagnetic field strength in the past is key to understanding deep Earth dynamics and detecting potential geodynamo regimes throughout the history of Earth. To better constrain the predictive power of the paleomagnetic record, we propose an approach based on the analysis of the dependency between geomagnetic field strength and inclination (angle made by the horizontal with the field lines). Based on the outcomes of statistical field models, we show that these two quantities should correlate for a wide range of Earth-like magnetic fields, even with enhanced secular variation, persistent nonzonal components, and severe noise contamination. Focusing on the paleomagnetic record, we show that the correlation is not significant for the Brunhes polarity chron, what we ascribe to inadequate spatiotemporal sampling. In contrast, the correlation is significant for the 1 to 130 Ma interval, whereas it only marginally succeeds prior to 130 Ma when strict filters on both paleointensities and paleodirections are applied. As we cannot detect significant variations in the strength of the correlation over the 1 to 130 Ma interval, we conclude that the Cretaceous Normal Superchron may not be associated with enhanced dipolarity of the geodynamo. The strong correlation obtained prior to 130 Ma when strict filters are applied indicates that the ancient field may not be on average so different from the present-day field. If long-term fluctuations nevertheless existed, detecting potential geodynamo regimes during the Precambrian is currently impeded by the sparsity of high-quality data passing strict filters in both paleointensities and paleodirections.