A Determination of the Rate of Change of G

Abstract

A new analysis of lunar occultation observations from 1955 to 1974 utilizing Atomic Time gives a value for the empirical part of the secular acceleration of the Moon's mean longitude of (− 65″ ± 18″)/cy². This differs from other determinations which utilized the Ephemeris Time scale, based on the apparent annual motion of the Sun about the Earth. These latter, which give (− 38″ ± 4″)/cy², measure only the tidal component of the Moon's longitude acceleration. The remaining acceleration (− 27″ ± 18″)/cy², has as its most probable cause a decrease in the Universal Gravitational Constant at the rate of $$\dot G/G$$ = (− 8 ± 5)× 10⁻¹¹/yr. There is a large body of supporting evidence for a rate of about this size, including the disappearance of the artificial ‘non-tidal’ component in the deceleration of the Earth's rotation, geophysical evidence for an observed expansion of the Earth's radius, and generally good agreement with recent determination of the Hubble constant, which is related to $$\dot G/G$$. The observed rate is also consistent with the Dirac and the Hoyle–Narlikar cosmological theories, and to a lesser degree, with the Brans–Dicke theory. Other possible interpretations of the observed excess (negative) lunar acceleration are also discussed; but only those with cosmological significance seem at all plausible.