Enhancing Astronomical Signal Detection in Cyclostratigraphy by Superimposed Averaging of Paleoclimate Proxies

Detecting allogenic signals like Milankovitch cycles in terrestrial and shallow marine strata is challenging due to their interplay with autogenic dynamics. To address this challenge, we proposed a superimposed averaging method for enhance signal detection based on the Dynamic Time Warping algorithm. This method improved the signal-to-noise ratio by accentuating periodic signals while suppressing stochastic noises in astronomical analyses. We first assessed the method by using synthetic series generated from astronomical solution. We found that superimposed averaging of two series significantly reduced the intensity of white noise by ∼27% and red noise by ∼17%. Further evaluation with Monte Carlo simulations showed that the method could significantly reduce non-Milankovitch variances after orbital tuning. The application of this innovative approach in the Upper Paleozoic Fengcheng Formation in Junggar Basin showed consistent noise reduction, further confirming the robustness of this method. We suggest that this method can support the study of astrochronology and astronomical forcing of paleoclimate evolution in dynamic depositional settings. Additionally, it can aid in researching the sedimentary processes influenced by both astronomical and non-astronomical factors across depressions, basins, or on a global scale.