交叉小波分析:检验日长振荡与D-O气候事件的相位是否真的锁定

Written by

in

Language

Evidence level: B

日长存在约1500年振荡,冰芯记录着1470年的丹斯果-奥什格尔事件——两个周期如此接近,究竟是共享同一个”节拍器”,还是纯属巧合?交叉小波相位分析是回答这个问题的关键实验。

正文

一、问题升级:从”周期接近”到”相位锁定”

上一篇文章论证了地球自转存在约1500年振荡(振幅约4毫秒,基于2700年日食与月掩星记录,PMID 28119545),而末次冰期的丹斯果-奥什格尔(D-O)事件以约1470年准周期重现(PMID 16281042)。周期接近只是”看起来像”,科学上需要更严格的判别标准:相位是否锁定。

被同一节拍器驱动的两个摆,即使周期不完全相同,相位关系也保持稳定;没有共同节拍器的两个独立振荡,相位会随时间随机游走。这条判别规则把问题变成一个可检验的假说:如果日长振荡与D-O事件共享外部驱动(如太阳),二者在相应频段应表现出稳定的相位关系;如果各自独立(日长源于核幔耦合,D-O源于气候系统内部振荡),相位角应呈随机分布。

二、方法学:交叉小波与相位角统计

检验工具是交叉小波变换(XWT)及其扩展——小波相干(WTC)。与传统傅里叶分析不同,小波方法能在”时间-频率”两个维度同时刻画两个非平稳序列的关系:相干性高的时频区域给出两序列同频振动的强度,相位箭头则给出局部的相对滞后方向。Boers 对 NGRIP 冰芯 5 年分辨率 δ¹⁸O 序列的小波分析提供了一个范本:D-O 事件的早期预警信号(方差与自相关增加)集中在 10-50 年频段,其显著性通过相位随机化检验(phase-randomization test)确认(PMID 29967370)——这套显著性框架正是相位角检验可以直接复用的。

相位角的统计显著性是关键环节。对相干区域内的相位角做圆形统计(circular statistics),检验其是否显著偏离均匀分布:若相位角聚集在某一方向且随时间稳定,即为相位锁定;若弥散,则排除直接耦合。蒙特卡洛方法(对原序列施加 AR(1) 或相位随机化扰动,重复上千次)用于估计置信区间——与 Boers 2018 检验早期预警信号显著性时使用的策略一致(PMID 29967370)。

三、真正的困难:时间窗口不重叠

这里有一个必须诚实面对的结构性困难:日长序列由古代日食/月掩星记录重建,只能回溯到公元前720年,覆盖约2700年(PMID 28119545);而D-O事件主要发生在末次冰期(约11.5万至1万年前)。两者在时间上不重叠——直接用XWT检验二者的相位,相当于把日长振荡外推到冰期,这超出观测数据的范围。

因此检验需要分两步走:

  1. 全新世替代检验(可立即执行):日长序列覆盖的2700年恰好落在全新世内,而全新世北大西洋存在千年尺度的漂冰碎屑(IRD)循环(Bond事件),且与太阳活动代理(¹⁴C、¹⁰Be)密切相关(PMID 11739949)。用XWT检验”日长振荡 vs Bond事件/太阳代理”在1500年频段的相干性与相位角,是真正可以马上做的第一步。
  2. 冰期频谱与相位结构比较(间接路径):对冰期D-O记录做频谱与相位统计——两极之间的相位差已被精确到百年以内(PMID 25925479)——再与日长振荡的相位行为做结构比较,判断两类记录的相位组织是否同类。

四、对解释的约束:D-O 可能根本不是”外部驱动”

相位检验的成败还取决于D-O事件本身的机制。Braun 的耦合模型证明,用约87年+约210年两个已知太阳周期叠加驱动非线性气候系统,就能自发产生1470年的响应(PMID 16281042)——这意味着”外部驱动”并不要求存在一个真实的1470年太阳周期,短周期叠加即可。更激进的假说是:1470年根本不是一个”真实”周期,而是幽灵随机共振的产物——系统在噪声与弱周期强迫共同作用下表现出的表观周期(PMID 20059228)。Crucifix 则系统梳理了更新世气候的振荡器与弛豫现象,指出D-O式事件更接近弛豫振荡而非外驱周期振荡(PMID 22291227)。如果这些假说成立,相位锁定的预期本身就要改写:弛豫振荡的相位会频繁漂移、被噪声重新设定,”稳定相位锁定”反而可能恰恰意味着存在外部驱动。

五、对易学时间观的启示

易学的时间假设是”天人合一、同频共振”。相位锁定检验提供了一个残酷而诚实的判别器:如果日长与气候的千年周期真的相位锁定,说明”天地同节律”存在可观测的物理基础;如果相位随机漂移,则周期接近只是巧合——”看起来同步”与”真的同步”之间,隔着一个相位角。这正是科学易学要做的事:不满足于周期数字的巧合,而是用可检验的方法区分巧合与耦合。

证据与局限

已验证: ①日长存在约1500年振荡,振幅约4毫秒,基于2700年记录(PMID 28119545);②D-O事件以1470年准周期重现(PMID 16281042);③全新世太阳活动与北大西洋IRD循环千年相关(PMID 11739949);④D-O事件两极相位差精确到百年内(PMID 25925479);⑤小波分析+相位随机化检验在NGRIP冰芯上的成功应用(PMID 29967370)

部分验证: ①87+210年太阳周期叠加可产生1470年响应(PMID 16281042);②D-O机制可能是弛豫振荡或幽灵随机共振而非外驱周期(PMID 22291227, 20059228);③”日长振荡与D-O周期相位锁定”的直接检验尚未见报道

待验证: ①全新世”日长 vs Bond事件/太阳代理”的交叉小波相干性与相位角检验;②冰期D-O记录与日长振荡相位结构的间接比较;③若相位锁定成立,需判别共同外部驱动(太阳)与内部反馈(核幔-气候)的具体路径

相关阅读

证据等级: B
参考文献: PMID 28119545, PMID 29967370, PMID 16281042, PMID 11739949, PMID 25925479, PMID 22291227, PMID 20059228


免责说明:本文属传统知识与科学文献的整理、对照与研究设计讨论,仅供理解与参考,不构成医疗建议,不能替代专业诊断与治疗。文中涉及的效应量、频率区间与暴露结论多为研究报告层面的推断,个体情况差异很大;如有身体不适或健康疑虑,请就医并由专业医师判断。

Cross-Wavelet Analysis: Testing Whether the Phase Between the Quasi-Biennial Oscillation and D-O Climate Events Is Truly Locked

Evidence level: B

Day length exhibits an oscillation of approximately 1500 years; ice core records capture the Dansgaard-Oeschger event at ~1470 years. Two cycles so close: do they share a single “metronome,” or is it mere coincidence? Cross-wavelet phase analysis serves as the critical experimental test for this question.

Main Text

I. Problem Escalation: From “Cycle Proximity” to “Phase Locking”

The previous article argued that Earth’s rotation exhibits an oscillation of approximately 1500 years (amplitude ~4 ms, based on 2700 years of eclipse and lunar occultation records, PMID 28119545), while the Last Glacial Maximum Dansgaard-Oeschger (D-O) event recurs with a quasi-period of ~1470 years (PMID 16281042). Cycle proximity is merely “appearing similar”; scientifically, a stricter discriminant standard is required: phase locking.

Two pendulums driven by the same metronome maintain a stable phase relationship even if their periods are not identical; two independent oscillations without a common driver exhibit random walks in phase over time. This discriminant rule transforms the problem into a testable hypothesis: if day length oscillations and D-O events share an external driver (e.g., the Sun), they should exhibit a stable phase relationship in the corresponding frequency band; if independent (day length arising from core-mantle coupling, D-O from internal climate system oscillations), the phase angle should be randomly distributed.

II. Methodology: Cross-Wavelets and Phase Angle Statistics

The testing tool is the Cross-Wavelet Transform (XWT) and its extension—Wavelet Coherence (WTC). Unlike traditional Fourier analysis, wavelet methods characterize the relationship between two non-stationary sequences simultaneously across both “time-frequency” dimensions: regions of high coherence indicate the intensity of same-frequency vibrations between two sequences, while phase arrows indicate local relative lag directions. Boers’ wavelet analysis of the 5-year resolution δ¹⁸O sequence from the NGRIP ice core provides a model: early warning signals for the D-O event (increased variance and autocorrelation) concentrate in the 10-50 year band, with significance confirmed via a phase-randomization test (PMID 29967370)—this significance framework is directly reusable for phase angle testing.

Statistical significance of the phase angle is a critical step. Circular statistics are applied to phase angles within coherent regions to test whether they significantly deviate from a uniform distribution: if phase angles cluster in a specific direction and remain stable over time, it indicates phase locking; if dispersed, direct coupling is excluded. Monte Carlo methods (applying AR(1) or phase-randomization perturbations to the original sequence, repeated thousands of times) are used to estimate confidence intervals—consistent with the strategy employed by Boers in 2018 when testing the significance of early warning signals (PMID 29967370).

III. The Real Difficulty: Non-Overlapping Time Windows

There is a structural difficulty that must be faced honestly here: the day length sequence is reconstructed from ancient eclipse/occultation records, retrievable only back to 720 BCE, covering approximately 2700 years (PMID 28119545); the D-O event primarily occurred during the Last Glacial Maximum (~115,000 to 10,000 years ago). The two are non-overlapping in time—directly applying XWT to test their phase relationship would require extrapolating day length oscillations into the glacial period, exceeding the range of observational data.

Therefore, testing requires a two-step approach:

  1. Holocene Substitution Test (Immediately Executable): The 2700-year day length sequence falls entirely within the Holocene, during which the North Atlantic exhibits millennial-scale Ice-Rafted Debris (IRD) cycles (Bond events) closely related to solar activity proxies (¹⁴C, ¹⁰Be) (PMID 11739949). Testing “Day Length Oscillation vs. Bond Events/Solar Proxies” for coherence and phase angles in the 1500-year band via XWT constitutes a genuine first step that can be performed immediately.
  2. Glacial Spectral and Phase Structure Comparison (Indirect Path): Perform spectral and phase statistical analysis on glacial D-O records—the phase difference between poles has been precise to within a century (PMID 25925479)—then compare the phase organization of these records structurally with day length oscillations to determine if they belong to the same class.

IV. Constraints on Interpretation: D-O May Not Be “Externally Driven”

The success or failure of phase testing also depends on the mechanism of the D-O event itself. Braun’s coupled model demonstrates that superimposing two known solar cycles (~87 years + ~210 years) can spontaneously generate a 1470-year response in the nonlinear climate system (PMID 16281042)—meaning “external driving” does not require a real 1470-year solar cycle; short-period superposition suffices. A more radical hypothesis is that the 1470-year cycle is not a “real” period but a product of ghost stochastic resonance—an apparent periodicity exhibited by the system under the combined action of noise and weak periodic forcing (PMID 20059228). Crucifix systematically reviewed oscillators and relaxation phenomena in Quaternary climates, noting that D-O-like events are closer to relaxation oscillations than externally driven periodic oscillations (PMID 22291227). If these hypotheses hold, the expectation of phase locking itself must be rewritten: relaxation oscillations exhibit frequent phase drifts and re-setting by noise; “stable phase locking” might actually imply the presence of external driving.

V. Implications for Yijing Time Views

The Yijing time assumption posits “unity of Heaven and Man, resonating at the same frequency.” Phase locking testing provides a harsh yet honest discriminator: if day length and climate millennial cycles are truly phase-locked, it indicates an observable physical basis for “heaven and earth sharing a rhythm”; if phases drift randomly, cycle proximity is merely coincidence—”appearing synchronized” and “truly synchronized” are separated by a phase angle. This is precisely what scientific Yijing Studies should do: not be satisfied with coincidental cycle numbers, but use testable methods to distinguish coincidence from coupling.

Evidence and Limitations

Verified: ① Day length exhibits ~1500-year oscillation, amplitude ~4 ms, based on 2700-year records (PMID 28119545); ② D-O event recurs with ~1470-year quasi-period (PMID 16281042); ③ Holocene solar activity and North Atlantic IRD cycles show millennial correlation (PMID 11739949); ④ D-O event shows polar phase difference precise to within a century (PMID 25925479); ⑤ Successful application of wavelet analysis + phase-randomization testing on NGRIP ice core (PMID 29967370)

Partially Verified: ① Superposition of ~87 + ~210 year solar cycles can generate 1470-year response (PMID 16281042); ② D-O mechanism may be relaxation oscillation or ghost stochastic resonance rather than externally driven periodicity (PMID 22291227, 20059228); ③ Direct testing of “Day Length Oscillation vs. D-O Phase Locking” has not yet been reported

Pending Verification: ① Holocene “Day Length vs. Bond Events/Solar Proxies” cross-wavelet coherence and phase angle testing; ② Indirect comparison of glacial D-O records with day length oscillation phase structure; ③ If phase locking is confirmed, need to discriminate specific pathways between common external driver (Sun) and internal feedback (core-mantle-climate)

Related Reading

Evidence Level: B
References: PMID 28119545, PMID 29967370, PMID 16281042, PMID 11739949, PMID 25925479, PMID 22291227, PMID 20059228


Disclaimer: This article compiles traditional knowledge and scientific literature for comparison and research-design discussion only. It is not medical advice and cannot replace professional diagnosis or treatment. Effect sizes, frequency ranges and exposure conclusions cited here are largely inferences at the research-report level, and individual circumstances vary widely. If you have any discomfort or health concerns, please consult a physician.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *