On forcings of length of day changes: From 9-day to 18.6-year oscillations

Abstract

We analyze fluctuations in Earth's rotation velocity (therefore also the equivalent length of day), using >50 years of IERS observations and the powerful method of Singular Spectral Analysis. The first 16 eigenvalues uncovered by SSA correspond to 10 components, all with physical sense. The first component is the trend, the second is the lunar node tide (18.6 yr, amplitude 1.3 ms). Next are variations with a period that implies forcing related to solar activity (11 years, 0.46 ms amplitude). Then, zonal oscillations linked to the solar (1 year, 0.81 ms; 0.5 year, 0.76 ms;) and lunar (27.54 days, 0.39 ms; 13.66 days, 0.73 ms; 13.63 days, 0.27 ms; 9.13 days, 0.14 ms) tidal potentials. The QBO at 2.36 years (0.08 ms) is interpreted as a Sun-related oscillation. The components at 13.63 and 13.66 days could contain a solar contribution. SSA is an efficient detrending algorithm and way to identify irregular (quasi-periodical) oscillatory components: its application to l.o.d. data yields refined observations, in good agreement with recent models but with some new results. There is no extracted component that could not be attributed to an existing periodic or quasi-periodic physical phenomenon. Progress in computers and signal processing has allowed us to resume and extend the analysis of l.o.d., and in particular to show evidence of a solar signature in the series. Uncovering the mechanisms through which solar activity acts on Earth's rotation is an exciting project for geophysicists that deserve renewed attention.