A Geomagnetic Estimate of Heliospheric Modulation Potential Over the Last 175 Years

Abstract Galactic cosmic rays (GCRs) interact with the Earth's atmosphere to produce energetic neutrons and cosmogenic radionuclides, such as 14C. The atmosphere is partially shielded from the interstellar GCR spectrum by both the geomagnetic and solar magnetic field. Solar shielding is often expressed as the heliospheric modulation potential, φ, which consolidates information about the strength and structure of the solar magnetic field into a single parameter. For the period 1951 to today, φ can be estimated from ground-based neutron monitor data. Prior to 1950, 14C in tree rings can be used to estimate φ, and hence the solar magnetic field, back centuries or millennia. Bridging the gap in the φ record is therefore of vital importance for long-term solar reconstructions. One method is to model φ using the sunspot number (SN) record. However, the SN record is only an indirect measure of the Sun's magnetic field, introducing uncertainty, and the record suffers from calibration issues. We here present a new reconstruction of φ based on geomagnetic data, which spans both the entire duration of the neutron monitor record and stretches back to 1845, providing a significant overlap with the 14C data. We first modify and test an existing model of φ based on a number of heliospheric parameters, namely the open solar flux, FS, the heliospheric current sheet tilt angle, α, and the global heliospheric magnetic polarity, p. This modified model is applied to recently updated geomagnetic estimates of FS and cyclic variations of α and p. This approach is shown to produce an annual φ estimate in excellent agreement with that obtained from neutron monitors over 1951-2023. It also suggests that ionisation chamber estimates of φ -- which have previously been used to extend the instrumental estimate back from 1951 to 1933 -- are not well calibrated. Comparison of the new geomagnetic φ with 14C estimates of φ suggests the long-term trend is overestimated in the most recent 14C data, possibly due to hemispheric differences in the Suess effect, related to the release of carbon by the burning of fossil fuels. We suggest that the new geomagnetic estimate of φ will provide an improved basis for future calibration of long-term solar activity reconstructions.