**1. Introduction**

Scholars have been studying the complex interactions between climate change and human history [1], and history is key to understanding the present and future. One of the major research themes of the Past Global Changes (PAGES) project focuses on the social impacts of historic extreme climate events and the responses, as well as the mechanisms and processes of past human-climate-ecosystem interactions at multiple spatial and temporal scales. It aims to enhance our understanding of the influence of contemporary climate change and the adaptation of human society. This is an international effort to coordinate and promote past global change research [2].

Among the studies on the impacts of and responses to the historic extreme climate events, most cases discussed the relationship between the human social system and the climate-environment system in the same region, but rarely involve regional correlations or common responses. In fact, when the impact of extreme climate events exceeds the regional carrying capacity, not only the affected areas, but also the initially non-affected areas can be influenced. There will be a common response in both affected areas and non-affected areas. For example, from 1813 to 1815, floods and droughts struck many countries of Europe, resulting in crop failures. Approximately 8000 refugees from Southern Germany migrated to Russia in the east. France, Italy, and the Netherlands imported food grains from Egypt, Russia, the United States, and some other regions [3]. In Australia, during the period

of 1800–1945, in the face of drought or floods, the social responses included the relocation of towns and the establishment of dams to coordinate water consumption in the upstream and downstream areas of the river basin [4]. In the southern part of North America, there were several severe drought events in the 9th to 14th centuries, and people there abandoned the infrastructures and migrated [5]. In China, from 1560 to 1890, it was at the height of the Little Ice Age that the climate fluctuated violently. The social responses, including famine-related migration and the allocation of money and grain, can be observed [6–12].

The presented case studies show that regional interaction has become an essential way of social response to historic extreme climate events. However, there is a lack of research on the characteristics, processes, and mechanisms of these regional interactions. Given the impact of extreme climate events and the social response could generate a complex multidirectional network in time and space [13], it is necessary to contribute to the research on the social response mechanism from the perspective of regional interaction.

In the present, global connectivity is continuously enhancing, and regional integration is deepening. It can be inferred that at the local, national, regional and global levels, the possibility of being affected by extreme events is increasing. According to the report of the American Meteorological Association, climate change is closely related to extreme events, and these events will seriously threaten the social economy and human life [14]. Therefore, inter-regional coordinated responses are urgently needed. IPCC's report pointed out that risk transfer and sharing will be an effective way of social response [15]; ye<sup>t</sup> considering the interdependencies between regional economic and social systems, it may have opposite effects on different regions, which means the disaster risks could be either reduced or even amplified for a certain region involved [16,17]. Although it is impossible to reproduce the exact results of the response to the past events, the mechanisms, experience, and lessons of regional interactions in response to the historic extreme climate events are still be of an essential reference value.

Using documentary evidence to study past extreme climate events has become a recognized method [18], which emphasizes China's advantages in researching the social impacts of and response mechanisms to past climate change. On the one hand, the monsoon climate in China is characterized by its instability, and the traditional agriculture-based economy made the socio-economic system significantly sensitive to the changes in climate. On the other hand, China owns abundant and continuous documentary records left by its long history, such as historical books, local chronicles, archived documents, private diaries, etc. Besides, newspapers, which were first published in China in the early 19th century, are the documentary records with a high temporal resolution. They can not only be used to reconstruct the precipitation, temperature and other weather conditions in history [19,20], but also to explore the whole development process of historic extreme climate events within the socio-economic systems [18,21].

In this paper, focusing on the famine-related migration and the allocation of money and grain, the spatial and temporal features of the regional interactions in response to the North China Famine of 1876–1879 (known in Chinese as the "Dingwu qihuang" or the "Incredible Famine of 1877–1878") are analyzed. Combining with exploration on the after-effects of the famine on both disaster areas and non-disaster areas, the results provide the empirical evidence for understanding social response mechanisms from the perspective of inter-regional linkages.

#### **2. Data Sources and Research Methods**

## *2.1. Case Selection*

North China is a region with a temperate monsoon climate, prone to drought in spring, summer and autumn. In the past 2000 years, there have been 227 extreme drought events in North China. Droughts occurred more frequently in 150–200 A.D., 550–800 A.D., 1050–1100 A.D., and 1850–1900 A.D. [22].

From 1876 to 1879, five provinces in North China, namely, Shanxi, Henan, Shaanxi, Hebei, and Shandong (Figure 1), suffered a severe drought. The reconstructed precipitation (wet/dry) series indicated that it was the most severe drought in this region in the past 300 years [23]. Sea surface temperature anomalies in the eastern Pacific region and intense El Niño events [24] had resulted in the weakening of the East Asian monsoon and precipitation variability, which were the direct causes of this drought event. It had global effects, with several regions experiencing extreme drought at the same time, including Australia, Europe, North America and South America [25–28].

**Figure 1.** Study area in the paper. In the North China Plain: Beijing, Hebei, Shandong, and Henan; In the Loess Plateau: Shanxi, Shaanxi, and Gansu; In the Yangtze River Basin: Sichuan, Guizhou, Hubei, Hunan, Anhui, Jiangxi; Jiangsu, and Zhejiang; andIn the Jiangnan region: Fujian, Guangxi and Guangdong.

In China, the year 1877 was classified as a "Ding" year, and 1878 was classified as a "Wu" year. Because the worst period was in 1877–78, this extreme drought event was historically known as the "Dingwu qihuang". In 1877, 20% of the villages in Shanxi Province experienced harvest failures, and in the central Shaanxi Province, the harvest rate of grains during the fall harvest season was merely 30%. The famine reached its peak in 1877, with Shanxi and Henan worst affected [29]. Worse still, an epidemic occurred soon after and had spread over a large area during the spring and summer of 1878.

Regarding the social factors, frequent warfare in the late Qing Dynasty, fiscal crisis, and overburdened tenants aggravated the severity of disaster [30–32], leading to severe damage to productivity, homelessness, and social crisis [23].

In the end, approximately 160 to 200 million people were affected by the drought, and about 9.5 to 13 million people died from famine and disease. Many worst-hit counties in Shanxi and Henan provinces had lost over 50% of their population, with the death toll passed 5 million and 1.8 million respectively [33].

## *2.2. Study Area*

The study area was divided into two parts (Figure 1). One is the a ffected areas. They are located mainly in the Loess Plateau and the North China Plain, which are the major wheat-growing areas in China, with a long development history and large population, containing the above-mentioned five drought-stricken provinces. The capital city and the political center of China, Beijing, is also in this region. This region is the target of the relief e fforts carried out by the Qing government. The Qing state's responses to the famine consisted of a variety of strategies, such as allocating relief silver and grain and reducing or canceling taxes.

The other part is the south region, containing the Yangtze River Basin and Jiangnan region. It is the place where the economic center of China in the Qing Dynasty was located, and the resources for the disaster relief mainly came from. The landforms of this region are featured with the plain area along the middle–lower reaches of the Yangtze River, and the hilly area in the southeastern part. Di fferent from the north region, this area is with a subtropical monsoon climate. Good hydrothermal conditions and well-developed water systems are conducive to the growth of rice, wheat, and a variety of cash crops, as well as the development of forestry and fishing.

## *2.3. Data Sources*

The data about the North China Famine of 1876–1879 were extracted from Qing Shi Lu [34], Shenbao [35], Disaster annals in recent China [36], and Qing Tong Jian [37] (Table 1).

Qing Shi Lu is a long-term compilation of the chronicles of the Qing dynasty. It contains a total of 4363 volumes. The materials in Qing Shi Lu are originally from the o fficial documents of the Imperial Cabinet and other departments, the pieces of writing from the National Historical Archives, and some first-hand materials such as the emperor's anthology and handwriting [10]. The historical materials in Qing Shi Lu are of exceptionally high value.

By 1876, the Shenbao had established itself as a commercially successful newspaper that carried the only public and serious discussion of many public issues in China [38]. From 1876–79, Shenbao's critical coverage of the famine focused not only on the five hardest-hit northern provinces but also on some other areas influenced by the drought event [39].

Disaster annals in recent China systematically and chronologically expounded on the natural disasters in China from 1840 to 1919, combined with explicit analyses on the time, location, extent, causes and social influences of various natural disasters, as well as the e ffectiveness and gaps of disaster mitigation measures [40]. The data is of high quality and reliability.

In addition to the above sources, data source about Dingwu qihuang is also supplemented by Qing Tong Jian. Its collection of historical materials is complete and reliable, and it discusses in detail politics, society, finance, economy, transportation, war, etc.


**Table 1.** Information on sources of the data about the North China Famine of 1876–1879.

After the removal of redundant records, a total of 186 historical records were extracted and classified (Table 2). We excerpted 96 records from Qing Shi Lu, 70 from Shenbao, 13 from Disaster annals in recent China, and 7 from Qing Tong Jian. Figure 2 shows that 1877–1878 is the key period, with significantly more records identified for the famine event.


**Table 2.** Classification of the historical records for the North China Famine of 1876–1879.


**Figure 2.** Percentage of days with records in each year from 1876–1879.

This paper selected data of wheat prices from the Food Price Database in the Qing Dynasty [41] (Table 1). The database is based on historical documents in the First Historical Archives of China and the Institute of Economics of the Chinese Academy of Social Sciences.

The data on the famine-struck area and plague-infested area is from Zhang's research [42] (Table 1), which was based on climate records extracted from historical documents in "A compendium of Chinese meteorological records of the last 3000 Years (in Chinese)" [43].
