Evidence level: B
把”火太过之年偏热、水太过之年偏寒偏涝”这句古典预报,拿到两千多个气象站、六十年的现代极端气候指数上去对账——对得上多少,就是对运气理论最好的量化体检。
正文
五运六气理论给每一个干支年份都贴了一张”气候预报标签”:年干逢甲己化土、乙庚化金、丙辛化水、丁壬化木、戊癸化火,阳干之年主”太过”,阴干之年主”不及”。火太过之年应当偏热,水太过之年应当偏寒偏湿,金太过之年应当偏燥——这套推演是《黄帝内经》运气七篇的骨架,也是两千年里医家们据以预测疫病、调整处方的底层时钟。问题在于:当现代气象学已经积累了足够的极端气候指数序列时,这套古典归年表到底经得起多大比例的对账?
要回答这个问题,第一步是把”太过不及”翻译成可检验的统计假设。检验对象不是某一年,而是把六十甲子按运气属性分组:火运太过组(戊年)、水运太过组(丙年)、金运太过组(庚年)……再检验各组在现代极端温度指数(TX90p 暖昼占比、TN90p 暖夜占比、TX10p/TN10p 冷昼冷夜占比)与极端降水指数(RX5day 最大五日降水量、R95p 强降水占比、CDD 连续干旱日数)上是否存在系统性差异。这正是前人做过的”干支-气象关联”研究的升级版:不再是单一城市、单一气象要素的个案,而是两千站以上、六十年以上的多中心大样本 meta 式对账。
已有文献给出了两条值得注意的证据线。一条线来自运气-气象-疾病直接关联研究:北京中医药大学团队用 1990-2004 年 15 年结核病数据与气象因子建模,发现发病与当年及前 1-3 年的气象因子显著相关,风力影响最强,且”前三年气象组合”优于单年数据,作者明确认为这验证了五运六气理论的实用性(PMID 26525546);同一团队对北京 1970-2004 年猩红热的复杂 Morlet 小波分析显示,其年发病率存在 5 年主周期,恰与五运的 5 年周期吻合,且高发年集中在金运之年(PMID 23713246)。这两项研究说明:干支归年确实在疾病序列中留下了可检测的周期印记。
另一条线是现代极端气候指数本身的可算性。2026 年发表的 ECHIDNA 数据集(Sci Data)基于 CMIP6 多模式集合,构建了中国主要流域极端气候历史与未来指数数据(PMID 42414340);2023 年 iScience 研究用偏差校正后的 CMIP6 模式评估了 1.5°C 与 2.0°C 温升下中国极端降水指数响应(PMID 37013188);内蒙古 1960-2017 年极端温度与降水事件时空变化研究(PMID 30172136)则展示了站级大样本分析的标准流程。这些文献共同表明:TX90p、RX5day 这类 ETCCDI 统一指数在中国已形成成熟的计算与质控体系,完全具备充当”检验标尺”的资格。
但把两条线接到一起时,必须正视一个关键难点——非平稳性。过去六十年中国极端暖事件显著增多、极端冷事件显著减少,这是全球变暖的确定性趋势(PMID 37013188、PMID 32755767)。如果直接拿”火太过年偏热”去对账而不剔除这个趋势,任何与暖化叠加的运气分组都可能出现虚假符合;反过来,若运气分组恰好与暖化趋势方向相反(如金运太过年集中在晚近的升温期),又会造成虚假不符。因此正确的检验设计必须做趋势校正:用去趋势距平、或用年分作为协变量回归,再检验干支分组的残差效应。这恰是”表观漂移与真实漂移分离”思路(见关联文章)在极端指数场景的延伸。
多中心大样本的 meta 设计还有第二层价值:它能把”运气符合度”按气候区、按海拔、按站距海岸远近分层报告。古典运气理论诞生于黄河中下游平原,它是否在西北干旱区、青藏高原区同样成立,本身就是悬而未决的问题。理想的分析应当输出一张”符合度热力图”:横轴是干支分组,纵轴是气候区,格子里是该组极端指数与理论预期的偏离方向与效应量——哪片区域站得住、哪片区域失灵,一目了然。
证据与局限
已验证: 干支年份存在可检测的疾病周期印记(猩红热 5 年主周期与五运吻合,PMID 23713246);运气-气象-疾病链条在多中心回顾数据上可建模且有效(PMID 26525546);中国站级极端气候指数计算体系成熟、数据充分(PMID 42414340、PMID 37013188、PMID 30172136)。
部分验证: “太过不及”的特定事件类型预测(火太过偏热、水太过偏寒偏湿)在个别城市气象序列上偶有吻合,但缺乏全国多中心、统一指数、去趋势后的系统对账;对运气周期与气候序列的相位关系(5 年/6 年)北京小波研究结论不一,部分研究未检出稳定周期。
待验证: 全国 2000+ 站极端指数按干支分组、去趋势后的效应量 meta;”符合度热力图”(气候区×干支分组);运气效应是否与 ENSO/PDO 等慢速气候振荡混杂或分离。
相关阅读
证据等级: B
参考文献: PMID 26525546, PMID 23713246, PMID 42414340, PMID 37013188, PMID 30172136, PMID 32755767
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。
Correlation Analysis Between the Excess and Deficiency of Heavenly Stems and Earthly Branches in Year Classification and Modern Extreme Climate Indices Based on Large-Sample Tes…
Evidence level: B
Translating the classical forecast—”years with excessive Fire tend to be hotter, years with excessive Water tend to be colder and wetter”—into a statistical comparison against over two thousand modern weather stations and six decades of extreme climate indices serves as the most rigorous quantitative health check for theories regarding fortune and timing.
Main Text
The Five Movements and Six Qi theory assigns a “climate forecast label” to every Gan-Zhi year: the heavenly stem determines the governing element (Jia/Ji transform to Earth, Yi/Geng to Metal, Bing/Xin to Water, Ding/Ren to Wood, Wu Gui to Fire); Yang stems indicate “excess,” while Yin stems indicate “deficiency.” Years with excessive Fire should be hotter, those with excessive Water colder and wetter, and those with excessive Metal drier. This deductive framework forms the backbone of the Seven Treatises on Movements and Qi in the Huangdi Neijing and serves as the underlying clock for physicians predicting epidemics and adjusting prescriptions over two millennia. The critical question remains: when modern meteorology has accumulated sufficient sequences of extreme climate indices, how robust is this classical year classification against statistical verification?
To address this, the first step is translating “excess and deficiency” into testable statistical hypotheses. The target of verification is not a single year, but grouping the sixty Jiabi cycles by their governing Qi attributes: the Excessive Fire group (Wu years), Excessive Water group (Bing years), Excessive Metal group (Geng years)… Then, verify whether these groups exhibit systematic differences in modern extreme temperature indices (TX90p warm day proportion, TN90p warm night proportion, TX10p/TN10p cold day/cold night proportions) and extreme precipitation indices (RX5day maximum 5-day rainfall, R95p heavy precipitation proportion, CDD consecutive dry days). This represents an upgraded version of previous “Gan-Zhi-Meteorology correlation” studies: moving beyond single-city, single-variable case studies to a multi-center large-sample meta-analysis spanning over two thousand stations and six decades.
Existing literature provides two notable lines of evidence. One line stems from research on direct links between Movements and Qi, meteorology, and disease: A team from Beijing University of Chinese Medicine modeled tuberculosis data from 1990–2004 against meteorological factors, finding significant correlations between incidence and meteorological factors from the current year and the preceding 1–3 years, with wind force having the strongest influence; notably, “three-year meteorological combinations” outperformed single-year data, and the authors explicitly concluded this validates the practical utility of the Five Movements and Six Qi theory (PMID 26525546). In a separate study by the same team, complex Morlet wavelet analysis of Beijing’s scarlet fever data from 1970–2004 revealed a dominant annual cycle of five years, aligning perfectly with the five-year cycle of the Five Movements, with peak incidence concentrated in Metal-governing years (PMID 23713246). These studies demonstrate that Gan-Zhi year classification leaves detectable periodic signatures within disease sequences.
The second line concerns the calculability of modern extreme climate indices themselves. The ECHIDNA dataset published in 2026 (Sci Data), based on CMIP6 multi-model ensembles, constructed historical and future extreme climate index data for major Chinese basins (PMID 42414340); a 2023 iScience study used bias-corrected CMIP6 models to assess responses of Chinese extreme precipitation indices under 1.5°C and 2.0°C warming scenarios (PMID 37013188); and a study on spatiotemporal changes in extreme temperature and precipitation events in Inner Mongolia from 1960–2017 (PMID 30172136) demonstrates the standard workflow for station-level large-sample analysis. Collectively, these literatures indicate that standardized indices such as TX90p and RX5day under the ETCCDI framework have established mature calculation and quality control systems in China, fully qualifying them as “verification scales.”
However, connecting these two lines requires confronting a key difficulty—non-stationarity. Over the past six decades, China has seen a significant increase in extreme warm events and a significant decrease in extreme cold events, a deterministic trend driven by global warming (PMID 37013188, PMID 32755767). If one directly compares “excessive Fire years being hotter” without removing this trend, any Qi grouping overlapping with warming may yield false positives; conversely, if a Qi grouping aligns opposite to the warming trend (e.g., excessive Metal years concentrated in recent warming periods), it may produce false negatives. Therefore, the correct verification design must incorporate trend correction: using detrended anomalies or regression with year as a covariate to test the residual effects of Gan-Zhi groupings. This extends the logic of “separating apparent drift from true drift” (see related article) to extreme index scenarios.
The multi-center large-sample meta-design also offers a second layer of value: it allows reporting of “Qi validity” stratified by climate zone, altitude, and distance to the coast. Born in the North China Plain along the Yellow River, the question remains whether the classical Qi theory holds true in arid Northwest regions or the Qinghai-Tibet Plateau. An ideal analysis should output a “validity heat map”: the horizontal axis represents Gan-Zhi groupings, the vertical axis represents climate zones, and each cell indicates the direction and magnitude of deviation between observed extreme indices and theoretical expectations—clearly showing which regions align with theory and which do not.
Evidence and Limitations
Verified: Detectable disease periodic signatures exist within Gan-Zhi years (the 5-year dominant cycle of scarlet fever aligns with the Five Movements, PMID 23713246); the Movements and Qi–meteorology–disease chain can be modeled effectively using multi-center retrospective data (PMID 26525546); China possesses a mature station-level extreme climate index calculation system with sufficient data (PMID 42414340, PMID 37013188, PMID 30172136).
Partially Verified: Predictions of specific event types based on “excess and deficiency” (e.g., excessive Fire being hotter, excessive Water being colder/wetter) show occasional alignment in individual city meteorological sequences but lack systematic verification across national multi-center data using unified indices after trend removal; conclusions regarding the phase relationship between Qi cycles and climate sequences (5-year/6-year) are inconsistent in Beijing’s wavelet studies, with some studies failing to detect stable cycles.
To Be Verified: Meta-analysis of effect sizes for extreme indices grouped by Gan-Zhi across 2000+ Chinese stations after trend removal; the “validity heat map” (Climate Zone × Gan-Zhi Grouping); whether Qi effects are confounded or separable from slow climate oscillations such as ENSO/PDO.
Related Reading
- Matching Test of Event Types in the Comprehensive Collection of China’s Three Thousand Years of Meteorological Records
- Validation of Gan-Zhi “Excess and Deficiency” Against Extreme Climate
- Research Pool: Environment-Life Frequency Interface
Evidence Level: B
References: PMID 26525546, PMID 23713246, PMID 42414340, PMID 37013188, PMID 30172136, PMID 32755767
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.
Leave a Reply