the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 19 Feb 2026
Analysis of the Relationship between Official Rain-Praying Rituals and Droughts in China over the Past 2000 Years
Abstract. Official rain-praying rituals, as an institutionalized cultural response to drought in ancient China, offer a crucial window into the evolution of state governance logic and disaster relief responses during climate crises. In this study, 1,825 official rain-praying records from the Western Han to the Qing dynasties were collated on the basis of the Twenty-Four Histories and the Qing Shigao, resulting in the construction of a long-term sequence with a ten-year resolution and a high-resolution annual sequence for the Ming and Qing periods. Sliding window correlation analysis was employed in combination with historical drought sequences to investigate long-term patterns of change. Functioning as a cultural disaster response indicator, this sequence, when integrated with other socioeconomic proxy indicators, enables a more comprehensive characterization of the climate change–impact–response process. The findings reveal that official rain-praying rituals exhibit a four-phase fluctuation pattern of low-high-low-high, with a significant nonlinear relationship between drought occurrence and such rituals. During the Han and Tang dynasties, rain-praying primarily constituted a direct response to environmental stress. In the Song and late Qing periods, confronted with crises of legitimacy amid internal turmoil and external threats, rulers favored high-frequency rain-praying to proclaim the mandate of heaven and pacify public sentiment, resulting in a strong correlation between rain-praying and drought. Conversely, during the Yuan and late Ming dynasties, influenced by ethnic cultural differences or the collapse of state administrative efficacy, a decoupling emerged whereby disasters occurred without corresponding rain-praying. Moreover, the Qing dynasty established systems of regular rain prayers and confidential memorials, transforming rain rituals from reactive disaster relief into proactive administrative routines. These rituals even exhibited a temporal lead over drought outbreaks, becoming institutionalized. During periods of relative fiscal abundance, state disaster responses prioritized substantive relief measures – such as opening granaries for distribution and tax reductions – over reliance on ceremonial rain prayers. As rulers' focus shifted from celestial principles to pragmatic grain storage for famine relief, the frequency of rain-praying rituals also declined markedly. Research findings indicate that ancient Chinese official rain-praying rituals were not determined solely by climate but constituted a complex political process regulated by state capacity, demands for political legitimacy, and institutional sophistication.
- Preprint
(1969 KB) - Metadata XML
-
Supplement
(114 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- CC1: 'Comment on egusphere-2026-187', Siying Chen, 09 Mar 2026
-
RC1: 'Comment on egusphere-2026-187', Anonymous Referee #1, 13 Mar 2026
This study compiles a long-term series of official rain-praying activities in China from 202 BCE to 1911 CE (1825 records) based on historical documents, and analyzes their correlation with historical drought records. It attempts to explain the evolution of disaster response behaviors from the perspectives of culture and political institutions. The article treats official rain-praying rituals as an indicator of social response and proposes that changes in their frequency were influenced not only by climatic conditions but also by factors such as state governance capacity, political legitimacy demands, and the degree of institutionalization. This research topic has a certain degree of novelty. Previous studies have often used rain-praying records as proxy indicators for climate reconstruction, while relatively few have systematically quantified them as social response behaviors themselves. The attempt to construct a long-term time series and interpret it within the framework of political history demonstrates considerable research potential. However, the current manuscript still presents several significant issues regarding data reliability and methodological rigor. Some key conclusions such as the “significant nonlinear relationship between drought occurrence and such rituals” remain largely at the level of narrative interpretation and lack rigorous quantitative testing. Therefore, I recommend major revision before the manuscript can be considered for publication.
1. The rain-praying data are mainly derived from sources such as The Twenty-Four Histories, Draft History of the Qing, and the Veritable Records of the Ming and Qing Dynasties. However, the recording density of these sources differs substantially. For example, the Ming and Qing Veritable Records are archival documents with near-daily entries, whereas earlier sources such as the dynastic histories of the Han and Tang periods are largely retrospective narrative compilations. This raises the possibility that rain-praying events in the Ming and Qing periods were more likely to be recorded than those in earlier periods. The manuscript briefly acknowledges this issue but does not conduct any quantitative correction. This potential bias needs to be addressed more rigorously.
2. Related to the issue above, I also have concerns about the comparability of the drought dataset. Although the authors use the authoritative source Collection of Meteorological Records in China for the Past 3000 Years, it is still possible that drought records from the Han to Yuan periods are underrepresented compared with the much richer documentary evidence from the Ming and Qing periods. This may lead to an underestimation of drought indicators in earlier periods. The authors are encouraged to incorporate additional proxy records reflecting drought variability for comparison and cross-validation in order to improve the reliability of the dataset.
3. Analytical methods are too simple to support complex historical interpretations; the core analytical methods of the paper are Spearman correlation and moving correlation analysis. However, the research questions involve complex issues such as political institutions, administrative capacity, social crises, and climate variability. Simple correlation analysis is insufficient to support the causal interpretations proposed in the manuscript. Many of the conclusions currently rely primarily on historical narrative rather than statistical testing. The manuscript repeatedly argues that increases in rain-praying frequency served functions such as demonstrating the Mandate of Heaven, stabilizing society, or responding to governance crises. However, these interpretations are supported only by historical examples rather than quantitative indicators. The authors are encouraged to introduce multivariate regression models to statistically examine the relationship between explanatory variables and the response variable. In particular, previously quantified indicators of social activity, such as war frequency, rebellion events, or fiscal stress, could be incorporated to test whether rain-praying activities indeed increased during periods of political crisis. Without such tests, these interpretations remain historical narratives rather than empirically validated conclusions.
4. The manuscript defines each record as one rain-praying event. However, historical sources may record the same event multiple times. A single rain-praying activity may last for several days, and actions in different locations may originate from the same administrative order. Therefore, the statistical unit of events may not be consistent. The authors should clarify whether duplicate counting exists and whether the time series may contain structural biases due to the counting method.
5. The authors propose that rain-praying activities exhibit a four-stage pattern (“low–high–low–high”). However, the division of stages appears to rely mainly on dynastic chronology rather than statistical testing. The stage classification should be based on time-series analysis and supported by statistical methods; otherwise, the segmentation may be subjective.
Citation: https://doi.org/10.5194/egusphere-2026-187-RC1 -
RC2: 'Comment on egusphere-2026-187', Anonymous Referee #2, 19 May 2026
General Comments:
This manuscript reconstructs the chronology of rain-praying rituals and conducts a correlation analysis between these rituals and the drought sequence, which makes a clear contribution to the study of the climate-society relationship. The amount of work involved in extracting relevant events from historical sources for quantification and the effort to categorize and further analyze all these records are impressive. The research itself is innovative as it takes rain-praying rituals as institutional and societal responses rather than purely climate proxies. The finding that historical “rain-praying rituals were not determined solely by climate but constituted a complex political process regulated by state capacity, demands for political legitimacy, and institutional sophistication” is valuable and implies great potential for publication and future studies. However, some issues need to be addressed here.
Specific comments:
1. The first comment is about the data extraction and quantification.
a) The frequency of rain-praying rituals was first constructed by conducting a lexical search using four keywords, but are those four keywords sufficient to cover all relevant events? In Chinese historical sources, many rain or drought-related ritual records use terms such as 祭 (sacrifice), 祀(worship), 求雨(praying for rain), 请雨(invoking rain), 禜or禳(apotropaic sacrifice / expiatory rite), 致祭(to present ritual offerings), and more.
There are plenty of examples like this across the dynastic histories of China, such as the one from 27 CE, 建武三年七月，雒阳大旱，帝至南郊求雨，即日雨. (As the author translated it in the Supplementary Table 1: “In July of the third year of Jianwu, Luoyang suffered a severe drought. The emperor went to the southern suburbs to pray for rain, and it rained that very day.”). This record is included, but the whole sentence does not contain any of these four keywords, so how these records were found and included in the primary search data should be further clarified.
(p.s., This should be translated as the seventh month, not July, as it refers to the lunar calendar. This is a problem in several records in the Supplementary Table. Please find and correct them as this can be misleading in terms of seasonality.)
Another example is the record from 1009 CE, which is also in the table but does not include any of the four keywords. From the current explanation, the secondary search is built on the results of the primary search with the four keywords, instead of supplementing it. Whether it is a methodological issue or a language issue, it needs to be further clarified.
b) The frequency of rain-praying rituals is mostly drawn from the dynastic histories, while many records from the Collection of Meteorological Records in China for the Past 3,000 Years are also from the same sources, especially before the Ming Dynasty, when local gazetteers were not yet widely used. As explained in Section 2.2.2, in the secondary search, many rituals were included only when drought or its relevant impacts were mentioned, which implies that there is a structural, non-independent coupling between the two sequences. For instance, the first record provided by the author in Table 3, “During droughts and major rain rituals, it is forbidden to light fires” (81 BC), is also the same record used to count the drought of this year in the 3,000-year records. There are lots of cases like this, which means that double encoding can significantly contribute to the correlation. It is necessary to discuss such circularity frankly, at least acknowledge such an issue, and soften the expression regarding causation.
c) It has been mentioned in the abstract and conclusion that there were 1,825 records of rain-praying rituals in total, but there were 1,835 in Table 1.
2. The interpretation of the correlation coefficients:
The statistical outcomes may imply different extents of correlation between the rituals and droughts in different historical periods, but this does not mean that the outcome can be directly explained by historical or political reasons. Statements regarding political legitimacy, state capacity, and administrative collapse are lacking independent evidence at the moment. For example, the paragraph from line 455 is a chunky section with many twists and direct attribution of frequent droughts and few rituals to multiple causation chains that require further proof. Sentences like “The Song Dynasty actively adjusted its policies, implementing practical measures such as constructing water conservancy projects while also emphasizing official rain-praying rituals to reassure the populace and demonstrate the Mandate of Heaven, particularly during confrontations with ethnic regimes such as the Liao and Jin. Influenced by Central Plains culture, ethnic regimes such as the Liao and Jin also adopted Song-style disaster response measures to consolidate their rule, developing hybrid official rain-praying rituals combining indigenous shamanic rituals with Sinicized ceremonies…” are strong claims but without any references or citations. This cannot be treated as common sense that all readers should know or be easily convinced of. If it is a sophisticated statement from other scholars, please cite them properly. If it is from the author’s primary research, some evidence from historical texts should be provided. The tone also needs to be softened, as those aspects are possible contributors to the outcome, but not necessarily the direct and only causes.
The same issues can be found in the next few paragraphs as well. For instance, the causation chain from the Mongols, to the elevated status of Shamanism and Buddhism, to the marginalization of Confucianism, and then to the neglect of agriculture is very complicated (if it existed) and should not be simply claimed as if it were common sense. Thus, further references and evidence are needed.
3. The sources, the citation of historical materials, and the historical factual errors.
a) The data on official rain-praying rituals were collected from the Chinese Digital Archives (https://www.zhonghuadiancang.com), which is not an authorized or widely recognized website. The problems with this type of website are that some historical characters are often omitted or replaced by similar characters, which is not good for lexical searches. I can understand that, as the website provides digital copies of dynastic histories, it is convenient to use. However, to make the research more solid, it is recommended that the author use more reliable sources (e.g., the national library) or at least examine the outcomes with those sources.
b) There are quite a few historical facts that need to be clarified.
For example, in line 588, during the severe drought in Zhili Province in the tenth year of Shunzhi's reign, Regent Prince Dorgon instructed… The tenth year of Shunzhi's reign is 1653. Dorgon died in 1650, so there may be an error in this direct quote. Also, for direct quotes like this, the proper way to cite such historical texts should clearly provide the name and volume/chapter of the source. It is also suggested that the authors should provide a supplementary document with the original Chinese text of these quotes and the titles of the cited works; otherwise, it is hard to trace or identify the original text from the current English translation.
Line 577, “Historical context supports this finding: as documented in Studies on Military Expenditures in the Qing Dynasty, following the Qing conquest of the Central Plains during the Shunzhi reign, the Southern Ming forces remained active, compelling the state to prioritize military expenditures—which consistently exceeded 70% of annual budgets.”
Such a statement involves a number, but lacks a clear source and citation, which needs to be clarified.
Line 534–535, 弘治中兴, the “Middle Revival,” usually refers only to the Hongzhi period. It may not be appropriate to use the term to cover the Zhengde reign. Line 733, “…during the Yongzheng reign, disasters were concealed to maintain a facade of stability”, is, again, a strong statement without evidence or references, especially since the reporting system in the Yongzheng reign is usually considered cautious. These statements can still be true, but evidence and references are needed.
There are also other small errors, such as the Chenghua era (1460–1487 AD). The Chenghua era started in 1465 AD. These should be corrected carefully.
Technical corrections:
Some expressions need to be fixed. Some sentences have grammatical mistakes.
For instance, in line 744, “potentially leading us overestimations” …
The mixed use of "Draft History of the Qing" and "Qing Shigao" can be confusing for international readers.
In the paper, especially in the section “Analysis of Types of Official Rain-Praying Rituals,” the names of different dynasties appear many times, which may confuse readers who are not familiar with Chinese history. I suggest adding the years of these dynasties at least right after they first appear.
In the abstract, “the Qing dynasty established systems of regular rain prayers,” the word “established” is a bit misleading. It is more like “regularized” or “advanced.”
Citation: https://doi.org/10.5194/egusphere-2026-187-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 1,344 | 618 | 82 | 2,044 | 106 | 92 | 167 |
- HTML: 1,344
- PDF: 618
- XML: 82
- Total: 2,044
- Supplement: 106
- BibTeX: 92
- EndNote: 167
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This paper investigates the long-term development of rain-praying rituals as a cultural response to climate extremes. The topic is both interesting and relatively underexplored, and the author’s attempt to address it is commendable. The study draws on a large body of historical documents and employs a combination of quantitative and qualitative analysis. The results contribute to a better understanding of how specific social and cultural systems shape decision-making and adaptation strategies in response to climatic stress. At the same time, several issues would benefit from further clarification and discussion.
(1) The manuscript frequently refers to Chinese dynastic periods as temporal anchors. For international readers unfamiliar with Chinese history, it would be helpful to provide the corresponding Gregorian years or centuries. This also applies to Figure 2.
(2) What does “era” refer to in Section 3.1.1? Does it correspond to a decade, a year, or another author-defined time unit?
(3) In the Results section, three different units are used to report the frequency of praying rituals: times/10a (e.g., lines 336 and 398), years/10a (e.g., lines 374 and 387), and records/10a (e.g., line 420). Why are these units mixed rather than using a single unit?
(4) Lines 445–454 suggest that during turbulent periods with limited human and material resources, the government tended to prioritize practical measures rather than rain-praying rituals. However, lines 614–619 state that when fiscal constraints prevented the implementation of measures such as conservancy projects and disaster relief, the government relied more on rain praying. These statements seem somewhat inconsistent.
(5) In Figure 4c, what do the three bars in different colors represent? The same question applies to Figure 5c.
(6) The paper mainly focuses on rain-praying activities initiated by the government. Were there also other forms of rain praying, such as those initiated by local communities or organized through religious institutions? A brief supplement would be helpful.