The available documentary data for historical climate reconstruction of China exceeds many other places within the Indian Ocean World. Archival materials, especially from official government records, have dominated investigations in this context. Most recently, a decades’-long effort led by Dr. Pao K. Wang resulted in the publication of the Reconstructed East Asian Climate Historical Encoded Series (REACHES) database, an extensive collection of environmental data between 1644-1795 based on written meteorological and climatical observations by Chinese government officials. At the beginning of 2023, the Indian Ocean World Centre at McGill University spoke to Dr. Wang about his research and this monumental effort. During the discussion, which was recorded as a podcast, Dr. Wang acknowledged that existing reconstructions of past environmental conditions in China based on historic documents are yet to account for all the available sources. Notable Western sources, such as those contained in missionary records, have yet to be adequately incorporated. As the REACHES database continues to be expanded, it is important to begin exploring these other sources to fill extant gaps. Following this suggestion, I began to study environmental data using the Mission Books of the Church Missionary Society (CMS) in China between 1834-1880. After gathering my data, I used QGIS to map climatic variability and possibly related environmental changes (see figure 1).
The first Mission Book begins in 1834 as the CMS explored the potential for “spheres of Missionary labour” along the Chinese coast. Between 1834-1840 the reports were limited, and the start of the Anglo-Sino War (First Opium War, 1839-42) hindered further exploration. However, following the signing of the Treaty of Nanking in 1842 at the war’s close, five Chinese ports were opened to Europeans, and Hong Kong was ceded to the British. This largely renewed the CMS’s interest in China, and in 1844, a post was established in Hong Kong. At the beginning of 1845, Ningbo and Shanghai were explored as sites of potential future missions, and stations appeared in both by 1847. In 1850, the CMS also established a post in Foochow. The Church Missionary Society focused on these centers along the Pacific and expanded into surrounding cities such as Hangchow in 1865. Finally, the CMS pushed north, and efforts can be found in Peking beginning in 1866. Other posts around these core locations did appear, but efforts and communications remained centered around these four primary locations. While stationed at their posts, CMS missionaries reported on a range of phenomena, including (but not limited to) their locales’ climatic, epidemiological, and economic conditions.
The Methodology
I tracked six themes in the Mission Books that indicated possible environmental changes—rainfall, famine, crops, disease, prices, and political stability.
I tracked six themes in the Mission Books that indicated possible environmental changes—rainfall, famine, crops, disease, prices, and political stability. Using human accounts of environmental data has distinct benefits and shortcomings. Firsthand accounts of occurrences such as disease, famine, and crop failures are largely reliable records that are important indicators of changes, be it by human or environmental actors. Similarly, missionaries frequently recorded specific quantitative data, such as number of patients monthly, which is unavailable from other sources. However, human actors are not impartial, and records reflect this.
In the Mission Books, there are far more accounts of heavy rainfall, wind, and extreme heat in the early 1840s than there are in later decades. As British missionaries were unaccustomed to the monsoon seasons of eastern China, it is difficult to conclude whether these years truly saw larger quantities of rainfall and more extreme heat, or if these accounts are a result of the acclimation of missionaries. However, when used in concert with other sources, such as the REACHES database, both predispositions and data itself may be verified. Therefore, when recording and indexing these themes, I did not analyze them critically. Rather, if a missionary wrote about “violent rainfall,” I recorded it this way to enable possible later comparisons, which can be the subject of future exploration.
Gathering Data
I initially gathered 330 lines of data from the nine Mission Books between 1834-1880. While I first tracked the specific locations of events, my findings reflect the generalization of these posts into four primary places: Macao/Canton/Hong Kong, Foochow, Ningbo/Hangchow/Shanghai, and Peking. Due to the vast amount of data, visualizing patterns at small, specific outposts proved difficult. Additionally, grouping stations together in this way allowed the visualization of contrasting patterns across all four locations. Further, Macao, Canton, and Hong Kong share the same water system, and today are often generalized together as the “Greater Bay Area.” Likewise, Ningbo, Hangchow, and Shanghai share the Hangchow Bay and proximity to the Qiantang River, which has important implications for localized environmental patterns.
Data itself varied, but frequently appeared similar to the following:
“The past month has been remarkably sickly and fatal. Mr. Lowder has had more funerals since his arrival than have taken place before since the port opened. This has been generally attributed to the overflowing of the river during the past summer when a great part of the city with the surrounding country and particularly the English land was entirely inundated.” (CMS CCHM1 Farmer to Lay Sec., 8 Nov. 1848, Shanghai)
Indexing the Data
This entry references the prominence of sickness and the flooding of rivers and was therefore entered as a data point for both rainfall and disease. I then indexed data by category, attempting to define my index values similarly to those in the REACHES database to ensure a degree of interoperability. Rainfall was indexed on a seven-point scale. A value of –3 indicates a lack of rainfall severe enough to result in famine. A value of 0 indicates normal rainfall, while a value of 3 indicates severe and heavy rainfall and flooding. Disease was indexed on a three-point scale. A value of 1 indicates a localized outbreak, a value of 2 an epidemic, and a value of 0 stipulates an uncertain level of transmission and/or mortality. In the previous example, rainfall was indexed as 3, and disease as 2.
Visualizing the Data
The visualization of data reveals several patterns, both between environmental factors and between sites themselves. On the map, the four central posts are highlighted in yellow and orange, with rainfall, crops, famine, and prices indicated to the right, and disease to the left. This decision was intentional. As the video plays, it is clear that rainfall is more highly correlated with the prominence of disease than any other factor. When above-average rainfall occurs, both the severity of disease and the number of diseases appear to increase. Once again, Reverend Farmer’s 1848 entry illustrates this point. Between October 1847 and March 1848, there are no mentions of disease anywhere in the Ningbo/Shanghai/Hangchow region, and only one account of average rainfall during this period. Beginning in April 1848, extremely heavy rainfall persists throughout the summer. Farmer describes a “remarkably sickly” season, which is indicated on the map from this entry. However, there are other records of diarrhea, fever, and ague during these dates. After the cessation of this heavy rainfall, accounts of disease largely disappear (see figure 2). There are significant exceptions however, such as 1872, in which there are accounts of extreme drought, poor harvest yield, and malaria, fever, typhus, eye diseases, and other unknown illnesses.
Further, disease appears to be more prevalent in the Ningbo/Shanghai/Hangchow region. This could be for several reasons. First, it is possible that proximity to both the Pacific and the Qiantang River made these areas more susceptible to disease. Wetter seasons would thus result in more water and marshier areas, which could increase the prominence of disease. Likewise, the number of large ports in this region could have made it vulnerable to the exchange of more diseases than Foochow and Macao/Canton/Hong Kong. However, it is also possible that there is simply more data in this region. As these three major CMS posts were in this area, there are more entries in the Mission Books. It is also known that British control of Hong Kong allowed missionaries to access more advantageous land at higher altitudes which was frequently considered healthier, as opposed to land missionaries were able to settle in other locations such as Ningbo/Shanghai/Hangchow.
However, neither above nor below-average rainfall seem to correlate as strongly with the strength of harvests, famine, or prices as it does with disease. While above-average to heavy rainfall appears to sometimes correlate with strong harvests, there are frequent exceptions, as illustrated in 1849 in Ningbo/Shanghai/Hangchow region and in 1876 and 1877 in Foochow. Similarly, famine often appears during poor harvests, but does not always. This suggests that famine during this period may have occurred because of other factors, such as political instability, warfare, or trade difficulties, rather than solely environmental changes.
Understanding Environmental Patterns
Environmental patterns across locations also appear in the data. Cholera is only mentioned a few times in the Mission Books, but when it appears in one location, it often does in another. In 1877, cholera was reported in the port of Shanghai. By September of that year, it was reported as a widespread catastrophe in Foo Chow. In both accounts, the missionaries describe that the most affected populations were the “natives” or impoverished. Similarly, despite differences in other environmental factors such as rainfall and harvests, particularly unhealthy years in one location appear to mirror those in others, perhaps demonstrating the interconnectedness of these systems and similar environmental phenomena. Famine patterns show this potential relationship as well. In 1877 and 1878, famine was reported in Peking, Foochow, and Tseng Un, present-day Taiwan. The entries credit the distress in all locales to the famine in the northern Chinese provinces, which was caused by a series of floods and droughts. These entries imply that the port cities largely relied on food from the north, and without it, suffered from lack of provisions. Thus, the interconnectedness of the posts is evident, but the precise impact of environmental factors on each location needs to be further explored.
This missionary data alone cannot provide conclusions for the relationship of the Hangchow Bay and disease, interconnectedness of trade and famine, nor necessary rainfall for high yield harvests. But it does provide a basis for these questions, with which other sources can be used collectively and analytically.
Of course, all this data has limitations. The frequency of letters varies significantly by year and location, and therefore the volume of data does not necessarily reflect real patterns. The Mission Books themselves are comprised of copied letters, not original papers. Therefore, it is possible some information has been lost, miscommunicated, or is missing altogether. During 1869 and 1870, the Mission Book summarizes the letters, rather than transcribing them. As environmental details are often found in the margins and additive details of these texts, this resulted in a drastic lack of evidence during those years. Some years, such as 1862-1868, are missing from the Mission Books altogether. Yet even still, this approach builds on the existing sources of environmental reconstruction, and itself provides a foundation to be built upon. This missionary data alone cannot provide conclusions for the relationship of the Hangchow Bay and disease, interconnectedness of trade and famine, nor necessary rainfall for high yield harvests. But it does provide a basis for these questions, with which other sources can be used collectively and analytically. Currently, the publicly accessible data from the REACHES database only extends between 1644-1795. Yet, as this becomes expanded, missionary and other records become salient sources for contributing to climate reconstruction and documenting environmental changes.
Feature Image: “File:View of Canton, artist unknown, China, 1800s, oil on canvas – Museu do Oriente – Lisbon, Portugal – DSC06833.JPG” by Daderot is marked with CC0 1.0.
Lilia Scudamore