Coherent interdecadal cycles in global rainfall, temperature, and cloud cover

Interdecadal cycles repeatedly appear in multiple global climate variables such as rainfall, temperature, and major climate modes, yet their origin remains a mystery and is most often attributed to quasi-periodic artefacts of internal climate variability or red-noise processes. Here we identify statistically significant (p < 0.001), coherent 12.9- and 19.9-year cycles in detrended rainfall, surface temperature, and total cloud fraction across ~40% of global land areas using the Gaussian clustering of wavelet amplitude power spectrum (GC-WAPS) method. GC-WAPS enables the aggregation of subtle cyclic signals across extensive spatial networks of climate records, providing robust discrimination from red-noise variability.

The resulting patterns exhibit organised regions of positive and negative phase alignment, forming large-scale teleconnection-like structures rather than isolated local responses. At sites exhibiting significant periodicity, the mean cycle amplitude in total cloud fraction is approximately 2%, corresponding to an estimated ~0.4°C modulation of surface temperature, consistent with estimates derived from longwave cloud radiative effect sensitivity. The cycles contributed an average of 10% to rainfall variance in significant regions, with the strongest signal being detected in eastern Australia, where the timing aligns with extended drought epochs. Regions of positive and negative correlation are mostly balanced, meaning these cycles represent a redistribution of energy rather than an overall heating or cooling effect.

The detected oscillations also align in period and phase with repeating gravitational cycles in Solar Inertial Motion (SIM), driven by the orbital dynamics of the Jovian planets, and consistently lag these dynamics by approximately two years. All five SIM periods shorter than 60 years correspond to cluster peaks in global rainfall within 1% error (12.9-, 19.9-, 29.4-, 35.9- and 45.4-years), though the three longer cycles are interpreted cautiously due to record length of each dataset. The observed phase coherence, amplitude, and cross-variable consistency motivate a tentative mechanism in which gravitational perturbations modulate interplanetary dust influx, leading to downstream changes in cloud cover, radiative balance, and rainfall; further work is required to test this hypothesis. These findings have implications for low-cloud feedback and decadal variability, highlighting a potentially externally timed component of Earth’s climate variability that is not explicitly represented in current CMIP-class models.