Transmission pathways of China’s historical climate change impacts based on a food security framework

Abstract

There are still many uncertainties about how climate change affects the development of human society. The impact of climate change is likely to be weakened or amplified by the response and adaptation of the reciprocal feedback process after entering the socio-economical subsystem. The study of the climate change impact processes is the basis for understanding the mechanisms of climate change impacts. Furthermore, long-term research of climate change impacts can provide historical similarity and experience for current or future adaptation of climate change. Ancient China was traditionally an agricultural country. Its food production safety, food supply safety, and food consumption safety reflect the impacts of climate change being transferred from a production subsystem to a population subsystem to an economic subsystem. Based on a conceptual framework of food security, we selected 10-year resolution sequences of grain harvest grades, famine indices, and economic levels in China over the past 2000 years (206 BC–AD 1911) to quantify and recognize the main transmission pathways of climate change impacts during cold and warm units of climate change. Our results were as follows: (1) according to the transmission relationship climate change → grain harvests → famines → economic levels, there are 16 main transmission pathways, including 10 pathways starting from cold units and 6 pathways starting from warm units. (2) The main pathway is dominated by natural factors and socio-economic factors, with each factor set accounting for approximately 50%. However, the probability of the main pathways with socio-economic influence leading toward negative developments was 60% for cold units, and the probability of the main pathways with socio-economic influence leading toward positive developments was 66.7% for warm units. According to the main pathway led by natural factors, of all transmission pathways in cold units (100%), 14.7% of the depressed economy may be more directly related to climate change. However, 32.3% of the prosperous economy may have a more direct relationship with climate change in warm units (100%). (3) Famine is the most important link in the climate change influence on transmission chains regulated by socio-economic factors. Famine reflects both poor harvests due to the natural production subsystem and a lack of grain allocation by the socio-economic subsystem or the government’s ability to eliminate famine. Thus, there may be another pathway of transmission, such as climate change → agricultural yield/grain production → economic prosperity → famine relief (social vulnerability).

Keywords

a food security framework ancient China from Western Han Dynasty to Qing Dynasty historical climate change impacts of climate change transmission pathway of climate change impacts

Proposing the problem

Social impact processes and mechanisms of climate change in historical periods are an important field in the current study of global warming and one of the research topics when studying past global changes (PAGES, 2014). History is the key to understanding the present and the future. The history of mankind’s experiences and lessons involved with climate change provide a valuable reference for modern societies and can help people to better understand current global climate change and actively respond to the challenges of future climate changes (IHOPE, 2010). In recent decades, there have been a large number of domestic and foreign studies that have revealed that climate change is one of the most important factors leading to the rise or fall of regional civilizations. Some social events are significantly synchronized with climate change or have close and complicated corresponding relationships: climate change could trigger a national war or population migration, affect economic fluctuation (Buckley et al., 2014; Butzer, 2012; Zhang et al., 2007), and even lead to civilization collapse (Haug et al., 2003). For example, in the history of Europe over the past 2500 years, the prosperity of the Roman Empire and the prosperity of the Middle Ages corresponded to warm periods of climate change, while the collapse of the western Rome Empire and the social upheaval during the Great Migration Period corresponded to periods of increasing climatic variability (250–600 BC; Büntgen et al., 2011; DeMenocal, 2001). Compared with case studies, studies on the influences of the processes and mechanisms of climate change are relatively lacking (PAGES, 2014).

China has a long and continuous history, documented by rich historical records that serve as a data source for the study of historical climate change having its own characteristics. Historically, China has had a large population and has been a largely agricultural country, with its population and agricultural activities being mainly concentrated in the eastern monsoon region. The instability of the monsoon climate seriously affects agricultural production, economic development, water conservancy facility construction, and people’s lives and safety. Therefore, there are many records in the literature associated with climate change and its impacts. Up until the present, the historical literature reconstruction of the existing social and economic indicators or sequences include the following details: grain harvest grades (Su et al., 2014; Yin et al., 2015), macroeconomic level indices (Wei et al., 2015a, 2015b), financial expenditure indices (Wei et al., 2014), social vicissitude indices (Yin et al., 2016a, 2016b), famine indices (Teng et al., 2014), peasant uprising indices (Fang et al., 2015), nation wars between agricultural and nomadic people indices (Bai and Kung, 2011; Su et al., 2016), and so on The temporal scale of these sequences are mostly covered the historical period of the Chinese monarchy (i.e. over the past 2000 years), and the data resolution is 10 years, which provides the foundational data for the quantitative study of the historical climate as well as its process and mechanism analyses. Based on the historical record and reconstructed sequences of socio-economical indicator, scholars have introduced quantitative research methods to discuss how climate change affects human society. It reveals the potential relationship between the climate change and the agriculture, population, economy, dynastic change, and social vicissitude in ancient China (Lee et al., 2008, 2010; Pei et al., 2015; Pei and Zhang, 2014; Xiao et al., 2015; Zhang et al., 2007, 2011a; Zheng et al., 2014).

To understand the mechanism of climate change impacts, it is necessary to clarify the processes of climate change impacts. Theoretically, the impact of climate change is transmitted along the three subsystems of the human system: the first is the environment and resource subsystem, then the support subsystem (including the production subsystem and infrastructure subsystem), and finally, the socio-economic subsystem (including population, economic and social subsystems; Fang et al., 2015; Wei et al., 2014). In other words, the transmission of climate change impacts is hierarchical. However, because of the vulnerabilities and resiliencies of human societies, response capabilities of climate change are very complex. Climate change impacts may be amplified, suppressed, or cut off in human systems through various reciprocal feedback effects (Ge et al., 2014). For example, previous studies have shown that famine is the most important link in the climate change influence on transmission chains regulated by socio-economic factors during the Chinese historical period. Obviously, the process of this connection and the intermediate links of climate change impacts need to be proven (Fang et al., 2015).

How are the impacts of climate change transmitted in social and economic system? Is the transmission pathway mainly influenced by natural factors or socio-economical factors? Through the discussion of these two issues, this paper attempts to reveal the climate change impact processes in ancient China. The results can provide historical similarity and experience for mankind to meet the challenges of global change and guarantee social economic sustainable development.

Data and methods

Conceptual design

According to the hierarchical relationships of climate change impacts and the characteristics of China as an agricultural country since ancient times, the historical impacts of climate change in China can be regarded as being related to food security issues, including the following three levels: food production safety (affected by natural system and food production subsystem), food supply safety (affected by population subsystem and economic subsystem), and food consumption safety (influenced by social subsystem; Figure 1). Historically, China has been an agriculture-oriented country. Climate change through affected food production subsystem to impact population subsystem and economic subsystem (Fang et al., 2014b, 2015; Su et al., 2014; Xiao et al., 2015). Specifically, the amount of food influences not only adequate food and clothing to impact the societal sensitivity to climate change but also the economic situation of the whole society to impact the response capacity of society to climate change (Yin et al., 2015, 2016b; Zhang et al., 2007; Zheng et al., 2014).

Figure 1.

A conceptual framework of the historical climate change impacts and responses based on food security.

Based on a conceptual framework of food security, we selected China’s eastern monsoon region as the study area. The sequences of temperature anomaly, grain harvest grades, famine indices, and economic levels were selected as indicators of climatic changes (the natural system), food production (the production subsystem), grain supply and demand (the population and economic subsystem), and economic prosperity (the social subsystem). Based on a conceptual model of food security, the main transmission pathway from climate change impacts to grain harvests, famines, and economic levels in each subsystem of Chinese society over the past 2000 years (extending from the Western Han to the Qing Dynasty, that is, from 206 BC to AD 1911) was analyzed.

Climate change is usually characterized by a change in temperature (cold or warm) and precipitation (dry or wet). Because of the relatively large regional differences and inter-annual variabilities of precipitation, the regional integrity reflected in the temperature is superior to precipitation (Fang et al., 2015; Pei et al., 2016). Therefore, the impacts of temperature changes are mainly studied in this paper. Grain harvests reflect relative fluctuations in grain yields, which are closely related to climate change and are the basis for climate change affecting food security. Therefore, in this paper, grain harvests were used as a relative variable to describe the food output capacity of the supporting subsystem. Famine is a phenomenon that a severe shortage of food resulting in violent starvation and death. Hence, the appearance of famine indicates a situation with food insecurity. The level of the economy indicates the level of the national wealth. China has been an agricultural country since ancient times, and its macroeconomic performance largely depends on whether there are bumper harvests or not. Since climate fluctuations have a significant impact on historical agricultural production, climate change can affect grain harvests and further influence the fluctuations of the whole macroeconomic subsystem. Therefore, in this paper, the grain harvests were used as a relative variable to describe the economic capacity of its socio-economic subsystem.

Date sources

Temperature anomaly sequence. Changes in temperature, especially in the winter temperature which is regulated by the Siberian high-pressure system, has a strong influence on crop growth, particularly during winter (Chu, 1973). The temperature anomaly (relative to the average winter temperature in 1951–1980) of the eastern winter half of the year over the past 2000 years (for October to the following April) recorded in the historical documents of the cold and warm phenological changes (such as plant phenology, the beginning and end of frost and snow events, the subtropical distribution of economic crops, agricultural planting, and harvesting dates) were selected; these data have a 30-year resolution (the resolutions of the two periods of AD 960–1109 and AD 1500–1999 can be up to 10 years; Ge, 2011; Ge et al., 2003; Figure 2). The sequence not only uses the temperature sequence of the past 5000 years from the historical phenological data reconstruction by Chu (1973) but also has all the advantages of reconstructing the temperature sequences from the historical literature as done in previous works, and removing any incomplete information. Meanwhile, the final sequence is highly correlated with other existing temperature reconstruction time series. This is the most authoritative and representative temperature sequence.

Grain harvest grade sequence. Yin et al. (2014) have reconstructed a 10-year resolution sequence of China’s grain harvest grade from 210 BC to AD 1910. Using the semantic differences method, the sequences were derived from the ‘imperial chronicles’ of every dynasty, Twenty-four Histories (Xu and An, 2004) and The Draft of Qing Dynasty (Zhao and Ke, 1977), combined with the 2755 annual crop yields; the conditions of food, agricultural disasters, people’s livelihoods; and food prices and storage records (Fang et al., 2014a). Thus, the semantic classification criteria of the implemented vocabularies were established. Then, the grain yield was graded and weighted according to the semantic differences and regional statuses of agricultural production in China before being assessed every 10 years. The sequence of the grain yields over the past 2000 years was built with a resolution of 10 years. In addition, in this sequence, the grain grades ranged from grade 1 (poor harvest) to grade 5 (bumper harvest) roughly equal to 44%, 56%, 64%, 71%, and 78% grain harvests, respectively.

Famine index sequence. Two sources are used to derive famine data. One source is the Compilation of Natural Disasters and Agricultural Policies in Dynastic China (Editorial committee of the Institute of History of the Chinese Academy of Social Science, 1988), which systematically lists all types of natural disasters occurring every year. ‘Hungry or Famine’ is recorded as a type of natural disaster events. The book recorded information about famine’s characteristics, including their times, places, victims, and so on. Another source is the History of Chinese Disasters (Yuan, 2008), which was divided into eight volumes to detail the disasters from the Qin Dynasty to the late Qing Dynasty. This book also lists the disaster chronologies within each dynasty and it extracted the ‘famine’ information of this chronology. There is a great difference in the detailed records of each dynasty. Therefore, the records of the famines were collected as two data sets, and a famine database was established. We have collected 4186 famine incidents. According to the statistical analyses, there are records of famines in 826 years from Western Han Dynasty to Qing Dynasty (a total of 2117 years), accounting for 39% of the total period. For more accurate quantitative analyses of the sequence of famines, two indices were calculated (refers to number of years with famine and frequency of famines experienced in China within per 10 years). Therefore, the creation of famine sequences can be used to reflect the frequencies of famines as well as the impacts of famines within the same dynasties. To eliminate the impact of this lack of homogeneity within historical document and to improve the precision of our famine sequence, we used a deviation standardization to regularize the number of years with famine and frequency of famines within per 10 years. First, according to the classification of the historical materials, the study period was divided into seven periods (the Western and Eastern Han; the Three Kingdoms, Wei, Jin and South-North; Sui and Tang; Song; Yuan; Ming and Qing). Second, the data were standardized via a function relation to formula (1):

X^{*} = \frac{X - \min}{\max - \min}

(1)

where X is the number of years with famines or the frequency of famines within 10 years; max and min represent the maximum and minimum values of this index in a certain time period; and X* is the standardized index, which has a value between 0 and 1.

Figure 2.

Sequences of temperatures, grain harvests, famines, and economic levels from 206 BC to AD 1911 in China: (a) winter half-year temperature anomaly (30-year resolution) in eastern China, (b) grain harvest grade (10-year resolution), (c) famine index (10-year resolution), and (d) economic level (10-year resolution).

Finally, we derive the ‘famine index’ through weighting by revising the two indexes. The index reflects the frequency and scope of the famine. The higher the number, the more the famines that occurred and the more widespread the affected area.

4. Economic level sequence. This sequence was derived from a 10-year resolution macroeconomic sequence for 220 BC–AD 1910 and was reconstructed from Wei et al. (2015b). The original data for this sequence were taken from 25 books on the history of China, including Chinese histories and economic histories written by contemporary well-known scholars. They collected 1091 records about the states of the macroeconomic level according to the semantic types of various vocabularies and their abilities to reflect macroeconomic fluctuations. The records were sorted into three groups with different levels, from high to low: macroeconomic status (458, 42%), macroeconomic process (576, 52.8%), and the farmers’ livelihood and national finances (57, 5.2%). In the same way, a semantic differential method was adopted to establish a semantic classification standard of the utilized vocabularies; this standard was related to the economic states of the previous dynasties. Then, they used a downscaling method integrated with multiresolution record data, reconstructed a 10-year resolution economic level sequence spanning from the Qin Dynasty to the Qing Dynasty (220 BC–AD 1910), and sorted the results into grades ranging from grade 1 (economic collapse) to grade 5 (economic climax).

Statistical analysis

Considering the resolution of famine and economic proxy indexes, cold units (<0°C) and warm units(>0°C) were delimited, according to the temperature anomaly of each 10-year unit (Table 1). Decadal grain harvests, famines, and economic levels were each classified based on three situations ranging from good to bad (Table 2). To analyze differences in social responses to climate change in cold and warm units, the decadal conditions of grain harvests, famines, and economic levels were calculated for each cold and warm units.

Table 1.

Classification of cold and warm units in China from 206 BC to AD 1911.

Climate types		Temperature anomaly	Centennial scale background	Notes
Cold units	Cold decades	<0°C	Cold periods	Centennial cold periods: (AD 180–570, AD 780–960, AD 1320–1910) Centennial warm periods: (200 BC–AD 180, AD 570–780, AD 960–1320)
Cold units	Cold decades in warm periods	⩽0°C	Warm periods
Warm units	Warm decades in cold periods	⩾0°C	Cold periods
Warm units	Warm decades	>0°C	Warm periods

Table 2.

Classification standards for each socio-economic index in China from 206 BC to AD 1911.

Indices	Situations	Classification standards	Classification basis
Grain harvest grade	Bumper harvest	>3	According to the definitions of grain harvest grades, 3 represents ‘normal’; 4 and 5 represent ‘near bumper’ and ‘bumper’, respectively; 2 and 1 represent ‘slightly poor’ and ‘poor’, respectively
	Normal harvest	3
	Poor harvest	<3
Famine index	Mild famine	<0.12	First, the threshold values are demarcated at the significantly different turning points of the famine index, which is also the frequency of low value points. Second, the normal distribution of the decades with the three famine situations is taken into consideration
	Moderate famine	0.12–0.45
	Severe famine	>0.45
Economic level	Prosperous economy	>3	According to the definition of economic classes, 3 represents ‘normal’; 4 and 5 represent ‘prosperity’ and ‘climax’, respectively; 2 and 1 represent ‘depression’ and ‘collapse’, respectively
	Normal economy	3
	Depressed economy	<3

To analyze the transmission processes of climate change in social subsystems according to the classifications shown in Tables 1 and 2, a dendrogram analysis was used to represent the layer-by-layer transmission processes. And it represents differences in various social subsystems for different types of cold and warm units. This analysis was based on hierarchical progressive relationships. Specifically, the proportion of the decadal experiences of each of the three grain harvest conditions (bumper, normal, and poor) was first calculated within the cold and warm units. Then, the proportions of each decade experiencing each of the three famine conditions (severe, moderate, and mild) were calculated for each temperature → grain harvest combination pattern. Finally, the proportions of the decades experiencing each of the three economic conditions (prosperous, normal, and depressed) were calculated for each temperature → grain harvest → famine combination pattern. In summary, this is the impact transmission pathway of climate change: temperature → grain harvest → famine → economy. Accordingly, the proportion of each impact transmission pathway was calculated. Based on the probabilities of pathway occurrences being greater than the average probabilities of the pathways, the main pathways were defined for different cold and warm units. Furthermore, different impact transmission pathways were categorized, and the possible influencing factors of the main pathways dominated by natural factors and socio-economic factors were analyzed.

Results and analysis

Possible stepwise transmission processes of climate changes

Table 3 shows the means of each subsystem index and the proportions of different situations in Chinese society during the different cold and warm units from the Western Han Dynasty to the Qing Dynasty. In all cold units, the proportion of bumper, normal, and poor harvest were roughly comparable, and the proportion of decades with moderate famines was the highest, accounting for only 42.6% of all years. In addition, the total combined occurrence percentage of severe and moderate famines was 75.4%. The total combined occurrence percentage of the normal and depressed economies was 75.6%, of which the proportion of decades with depressed economies was the highest, accounting for 44.3% of the total years. On the other hand, during the warm units, the total combined percentage of bumper and normal harvests was 81.1%, and thus, the number of years experiencing these conditions was significantly higher than the number of poor harvest years. The proportion of decades with severe famines was the lowest, accounting for only 24.4% of all years. In addition, the total combined percentage of mild and moderate famines was 75.6%, wherein the percentage of prosperous economic years was 56.7% and the percentage of depressed economic years was 20.0%.

Table 3.

Proportion distribution of different situations and mean values of each cold and warm units subsystem within China from 206 BC to AD 1911.

Periods	The percentage of the grain harvest scenarios (%) and mean grade		The percentage of famines degrees (%) and mean value of the index		The percentage of economic levels (%) and mean value of the index
Periods	Poor:normal:bumper	Mean	Severe:moderate:mild	Mean	Depressed:normal:prosperous	Mean
Total periods	26.9:35.8:37.3	3.14	29.3:40.1:30.6	0.31	34.0:28.8:37.2	3.07
Cold units	32.8:32.0:35.2	3	32.8:42.6:24.6	0.34	44.3:32.8:22.9	2.72
Warm units	18.9:41.1:40	3.31	24.4:36.7:38.9	0.28	20.0:23.3:56.7	3.55

In comparing the differences between the cold and warm units, we found that the history of China presents the characteristics of population increases and economic development during warm periods and population decreases and economic recession during cold periods. The proportions of decades with bumper and normal harvests during warm units were 1.13 times and 1.28 times greater than those for cold units, respectively. However, the proportion of decades with poor harvests during warm units was only 57.6% of the cold units. The proportions of decades with severe and moderate famines during cold units were 1.34 times and 1.16 times greater than those for warm units, respectively, whereas the proportion of decades with mild famines during cold units was only 63.2% of the warm units. The proportions of decades with depressed and normal economies during cold units were 2.21 times and 1.41 times greater than those for warm units, respectively. However, the proportion of decades with prosperous economies during cold units was only 40.4% of the warm units.

Based on the transmission relationship of climate change → grain harvests → famines → economic levels, Figure 3 shows the possible transmission pathways and the number of years wherein temperature change is transmitted to grain harvests, famine, and economic levels in China over the past 2000 years using dendrogram analysis. From the temperature level to the grain harvest level, bumper harvests and poor harvests could occur during both cold and warm units, but there are obvious differences in probabilities (Figure 3 and Table 3). Specifically, the average grade of the grain harvests for warm units was higher than the average grade for cold units.

Figure 3.

Different proportion in the possible transmission pathways of the climate impacts for the 10-year cold and warm units within the subsystems of Chinese society from 206 BC to AD 1911. (According to the idea of the stepwise transmission, the number from top to bottom is equal to 100.)

From the grain harvest level to the famine level, mild, moderate, and severe famines can occur in all kinds of situations, but the proportions differ (Figure 3). During a poor harvest, the proportions of mild, moderate, and severe famines were 11.7%, 41.2%, and 47.1%, respectively. For the normal harvest situation, these famine levels were 37.8%, 32.4%, and 29.7%, respectively. For the bumper harvest situation, these figures were 52.8%, 38.9%, and 8.3%, respectively. Therefore, as the grain harvests range from poor to bumper, the incidences of severe famine decrease gradually within warm units, while the incidences of mild famines increase. During cold units, for the poor harvest situation, the proportions of mild, moderate, and severe famines were 12.5%, 57.5%, and 30%, respectively. For the normal harvest situation, these figures were 30.1%, 48.7%, and 20.5%, respectively. For the bumper harvest situation, these figures were 30.2%, 23.2%, and 46.5%, respectively. Similar to the warm units, with the grain harvest ranging from poor to bumper, the incidence of mild famine increased gradually, but the incidence of severe famines was the lowest for the normal harvest situation. Note that, during cold units, the incidences of severe famine with bumper harvest situations were more frequent than those for poor harvests. These results indicate that, in addition to being affected by the grain harvest, famine was likely affected by other factors.

From the famine level to the economic level, prosperous, normal, and depressed economies could occur in all kinds of situations, but the proportions differ (Figure 3). In the warm units, for the cases of mild famines, the proportions of the prosperous, normal, and depressed economies were 57.1%, 17.1%, and 25.7%, respectively. For the moderate famine situations, these proportions were 54.5%, 27.3%, and 17.1%, respectively. For severe famine situations, these figures were 59.1%, 27.3%, and 13.6%, respectively. The probability of a prosperous economy was between 55% and 60% under all three famine types. During cold units, for mild famines, the proportions of the prosperous, normal, and depressed economies were 10%, 30%, and 60%, respectively. For the moderate famines, these figures were 19.2%, 34.6%, and 46.2%, respectively. For the severe famines, these figures were 32.5%, 37.5%, and 30%, respectively. The probability of a prosperous economy during cold units was less than half that during the warm units. This indicates that the accumulation of national wealth and the progress of economic strength are not only related to food security but also influenced by climate change. Specifically, during warm periods, there is a greater likelihood that people or countries will develop industries and commerce to promote their economic development as long as food security issues are resolved. However, during cold periods, although the social (agriculture) production was normal and provided people with adequate food and clothing, the probability of a prosperous economy was lower than that during warm units. In these periods, the society was not more prosperous, but the people lived and worked in peaceful situations, although the economy sometimes faced recession and crisis.

The main transmission pathways of climate change impacts during cold and warm units

Starting from cold or warm units and ending at the economy, there were 27 (a total of 54) possible transmission pathways for climate change → grain harvests → famines → economy pathways for each type of unit. The proportion of decades with cold or warm units was set to 100%. The average rate of every pathway was 3.7%. According to the statistics, 50 transmission pathways occurred in reality, of which 24 for warm units and 26 for cold units. The pathway probabilities greater than the average probability (3.7%) for the transmission pathway is defined as the main transmission pathway for 10 pathways starting from cold units and 6 pathways starting from warm units (Table 4).

Table 4.

The 16 main transmission pathways of climate change impacts in China from 206 BC to AD 1911.

Temperature units	Primary factor		Percentage (%)	Main possible pathways: Harvest → Famine → Economy
Cold	Natural factors
			4.1	Bumper → Mild → Normal
			5.7	Normal → Moderate → Normal
			5.7	Poor → Moderate → Normal
			9.8	Poor → Moderate → Depressed
			4.9	Poor → Moderate → Depressed
	Socio-economical factors	Negative progression	4.9	Bumper → Mild → Depressed
			7.4	Normal → Mild → Depressed
			7.4	Normal → Moderate → Depressed
		Positive progression	7.4	Bumper → Severe → Prosperous
		Positive progression	5.7	Bumper → Severe → Normal
Warm	Natural factors		15.6	Bumper → Mild → Prosperous
			10	Bumper → Moderate → Prosperous
			6.7	Normal → Mild → Prosperous
	Socio-economical factors	Negative progression	7.8	Normal → Mild → Depressed
		Positive progression	6.7	Normal → Moderate → Prosperous
		Positive progression	8.9	Normal → Severe → Prosperous

Main transmission pathways dominated by natural factors

Of the pathways starting from cold units, there were five transmission pathways with normal logical reasoning dominated by natural factors (Table 4): cold (100%) → bumper harvests (35.2%) → mild famines (10.7%) → normal economy (4.1%), cold (100%) → normal harvests (32.0%) → moderate famines (15.6%) → normal economy (5.7%), cold (100%) → poor harvests (32.8%) → moderate famines (18.9%) → normal economy (5.7%), cold (100%) → poor harvests (32.8%) → moderate famines (18.9%) → depressed economy (9.8%), and cold (100%) → poor harvests (32.8%) → severe famines (9.8%) → depressed economy (4.9%). Of all the transmission pathways during cold units (100%), 14.7% of depressed economy may have a more direct relationship with climate change.

Of the pathways starting from warm units, there were three transmission pathways dominated by natural factors: warm (100%) → bumper harvests (40%) → mild famines (21.1%) → prosperous economy (15.6%), warm (100%) → bumper harvests (40%) → moderate famines (15.6%) → prosperous economy (10%), and warm (100%) → normal harvests (41.1%) → mild famines (15.6%) → prosperous economy (6.7%). Of all transmission pathways in warm units (100%), 32.3% of prosperous economies may have a more direct relationship with climate change.

Main transmission pathways dominated by socio-economic factors

Of the pathways starting from cold units, there were five transmission pathways that were inconsistent with normal logical reasoning, and there were three transmission pathways that showed economic depression concurrent with bumper or normal grain harvests and the mild or moderate famines (Table 4): cold (100%) → bumper harvests (35.2%) → mild famines (10.7) → depressed economy (4.9%), cold (100%) → normal harvests (32.0%) → mild famines (9.8%) → depressed economy (7.4%), and cold (100%) → normal harvests (32.0%) → moderate famines (15.6%) → depressed economy (7.4%). In addition, these pathways indicate that the economic conditions of these decades are affected by climate change but are mainly affected by the interferences of other socio-economic factors, such as policy, war, social status, and other regulatory factors, which induced negative development. There were two transmission pathways that showed severe famines concurrent with bumper grain harvests wherein the severe famines did not further trigger economic depressions but instead induced positive developments. These two pathways are as follows: cold (100%) → bumper harvests (35.2%) → severe famines (16.4%) → prosperous economy (7.4%), cold (100%) → bumper harvests (35.2%) → severe famines (16.4%) → normal economy (5.7%). In such circumstances, the famine was more likely to indicate local food supply and demand imbalances in certain region, which can be controlled through government relief measures (relief, grain adjustment, work relief, etc.). Although the disasters were serious, the effects of the famine were not transferred to the economic level and the recorded relief of the famine reflected the government achievements. In addition, the transfer may also be related to the population, such that with a bumper grain harvest and prosperous economy, the population growth rate was high. If the grain harvest rate was lower than the population growth rate, there may be a decrease in the per capita grain consumption, which would lead to famine.

There were three main pathways starting in the warm units that were mainly dominated by socio-economic factors, only one of which showed negative developments: warm (100%) → normal harvests (41.1%) → mild famines (15.6%) → depressed economy (7.8%). There were two transmission pathways that showed moderate and severe famine concurrent with a normal grain harvest wherein the effects of the famines did not reach the economic level. Instead, these pathways led to positive developments and economic prosperity. The two pathways are as follows: warm (100%) → normal harvests (41.1%) → moderate famines (13.3%) → prosperous economy (6.7%) and warm (100%) → normal harvests (41.1%) → severe famines (12.2%) → prosperous economy (8.9%).

Discussion

The differences in the attributes of the main pathways before and after AD 1100

The statistical analyses reveal that AD 1100 was a demarcation point in China over the past 2000 years (Table 5). Before AD 1100 years, the main transmission pathway of climate change was dominated by natural factors, accounting for 59.5% of all main pathways. The main transmission pathways of climate change led by socio-economic factors mostly led toward negative developments. After AD 1100 years, the main transmission pathways of climate change were dominated by socio-economic factors, and most of these led toward positive developments. However, the proportion of the main pathways led by natural factors was only 31.8%.

Table 5.

The primary factors and the corresponding periods of the main transmission pathways affected by climate change from 206 BC to AD 1911.

Boundary	Primary factor		Percentage (%)		Corresponding periods
Before AD 1100	Natural factors		59.50		180–120 BC, 80–50 BC, 40–30 BC, AD 0–30, AD 90–100, AD 190–200, AD 250–280, AD 290–300, AD 310–320, AD 330–340, AD 350–390, AD 460–470, AD 490–500, AD 550–570, AD 590–600, AD 640–660, AD 670–680, AD 710–750, AD 780–810, AD 850–890, AD 920–930, AD 940–950, AD 960–1010, AD 1020–1060, AD 1080–1090
	Socio-economic factors	Positive progression	40.50	16.70	AD 70–90, AD 580–590, AD 660–670, AD 690–710, AD 950–960, AD 1010–1020, AD 1090–1100
	Socio-economic factors	Negative progression	40.50	23.80	100–90 BC, AD 30–60, AD 170–180, AD 200–250, AD 340–350, AD 390–430, AD 450–460, AD 520–530, AD 890–920, AD 930–940
After AD 1100	Natural factors		31.80		AD 1330–1350, AD 1450–1460, AD 1490–1500, AD 1600–1610, AD 1650–1660, AD 1860–1870, AD 1880–1910
	Socio-economic factors	Positive progression	68.20	40.90	AD 1140–1150, AD 1170–1190, AD 1200–1210, AD 1280–1300, AD 1400–1410, AD 1420–1430, AD 1680–1760, AD 1770–1790, AD 1800–1830
	Socio-economic factors	Negative progression	68.20	27.30	AD 1120–1130, AD 1240–1260, AD 1360–1380, AD 1460–1470, AD 1610–1620, AD 1840–1860

Through this analysis, the dominant pathways were found to differ before and after AD 1100. The reasons for this phenomenon may be related to the center of the economic development of China moving southward. In an agricultural society, climatic factors were an important determinant of the economic conditions. The southward move of China’s ancient economic development center was born at the Tang Dynasty (AD 618–907) and Five Dynasties (AD 907–960) and was basically completed by the Southern Song Dynasty (AD 1127–1279). By the end of the South Song Dynasty, the center of economic development in China had shifted completely to the south, and this moment expanded the scope of governance as well as the separation of the southern economic center and the northern political center. The results were that the economy increased the north’s dependence on the south, promoting the connection between the south and the north. At the same time, the opening of canals and other routes increased the contact between the north and south, reducing social vulnerability as well. Meanwhile, due to the changes in the supply relationship between the food production (South) and food distribution systems (North), the cooperation between the two sides of the food supply was more controlled by socio-economic factors, showing that natural factors (climate change, natural disasters, etc.) had gradually decreasing influences and that social systems had gradually increasing influences on climate change.

In all the main transmission pathways identified, there were 31 periods characterized by mild or moderate famines → depressed economies; these periods experienced negative developments. Among these 31 periods, 10 occurred during the periods of dynastic alternations or regime divisions. These were the late Eastern Han Dynasty – the Three Kingdoms period (AD 220–250), the Sixteen Kingdoms – Northern and Southern Dynasties period (AD 390–430), the late Tang Dynasty – Five Dynasties and Ten Kingdoms period (AD 890–920), the replacement period of the Southern Song Dynasty and the Northern Song Dynasty (AD 1127), and the replacement period of the Yuan Dynasty and Ming Dynasty (AD 1360–1370). Societal decay and the changes in the ruler causing the large-scale fighting and the upheaval of the political situation as well as the invasion of foreign countries and the division of state powers influenced the transmission pathways that led to negative development.

There were 30 periods that presented severe or moderate famines → prosperous economy; these periods experienced positive developments. Among these 30 periods, 20 occurred during times of peace and prosperity: the flourishing ages of the Eastern Han (AD 70–90), Sui (AD 580–590), Southern Song (AD 1140–1150), Yuan (AD 1280–1300), Ming (AD 1400–1410), and Qing (AD 1680–1760) periods. In these periods, the whole society was stable, the state government was unified, the economy was exempted from taxation, and agriculture was vigorously developed to achieve the prosperity of the country. Under these circumstances, in spite of there were local food shortages and hungry people civilians, the overall vulnerabilities of the society were reduced. The court implemented timely controlled interventions of social and government forces to reduce the severities of the famines and maintain the stability of the social system, reducing the effects of local famines. In this way, the social impacts of climate change were not transmitted to economic depressions, and the transmission pathways eventually led to positive developments. In addition, these cases presented peaceful, thriving, and flourishing societies.

Differences in event characteristics reflected in famine records

Famine is the most important link in the climate change impact on transmission chains regulated by socio-economic factors. Famine is affected by both the natural grain production subsystem and the population and economic subsystem, which has both natural and social attributes. Historical records of famines indicate that many famines were caused by natural disasters, such as droughts and floods. Therefore, there are famines caused by natural reasons found in the historical records, which reflect the food insecurities of the amount of grain per capita. The occurrences of famines were not only related to the variations in natural systems but also to socio-economic factors. Some famines were due to political and economic upheavals (such as war and the destruction of trade), official negligence, corruption, and so on.

In addition, some famine records indicated that local food shortages can be eliminated by societal and government measures to alleviate and eliminate hunger, such as grain production and circulation (food relief measures) and exemption from taxes. For this reason, there are some famine relief records in the historical documents. These types of famines do not lead to collapses of the population, but indirectly reflect an overall economic status of the society having the ability to implement relief measures at this time. Therefore, these records of famines represent the achievements of the government, which can be indirectly regarded as societal evidence of economic prosperity and great national strength. For example, in the Zhu Di period of the Ming Dynasty, although local areas still had famine refugee and vagrants, through a relief policy (migrating famine refugee to the food-producing areas or bringing food and other supplies from the fields to support the disaster areas), the famine was alleviated. Therefore, this historical period in the famine records mostly reflects a famine event that can be prevented by the central government taking relief and immigration measures.

Therefore, there are differences in the event characteristics reflected in the famine records. For the main pathways of the cold and warm units, the main pathways dominated by natural factors and those dominated by socio-economic factors account for 50% each. However, the probabilities of the main pathways of the cold units with social influences leading to negative developments was 60%, and the probability of the main pathways of the warm units with socio-economic influences leading to positive developments was 66.7%. During a relatively harvest period, the government has a stronger economic capacity. At this time, the recorded famines could be solved by government relief, food transportation, and other artificial regulation measures, allowing for positive developments.

Other feedback pathways for the transmission of climate change impacts

For the two attributes of famine, the influences of food security from the impacts of climate change in ancient Chinese societies could have two different transmission pathways. The first one is climate change → agricultural yield or grain production → the per capita supply of food → famine refugees (famines) → economic prosperity. In this pathway, temperature changes indicate natural system changes, grain harvests indicate the production capacities of the support subsystem, the natural attributes of the famines (the occurrence of famine) characterize the insecurity of the food supply and the stability of population subsystem, and the economic level indicates the overall socio-economic status of the country. The second pathway represents the vulnerability of the society in terms of the social attributes of famine, which indirectly indicates the degree of prosperity of the economy. Therefore, the key points of this transmission are as follows: climate change → agricultural yield or grain production → economic prosperity → famine relief (social vulnerability). The first transmission pathway is analyzed in this paper.

The definition of the transmission pathway climate change → the grain harvest → famine → economy was based on using the natural properties of the famine to characterize the stability of the population subsystem. While for the transmission pathway dominated by socio-economic factors, the socio-economic situation may be directly affected by the food production capacity of the supporting subsystem, which further affects the population growth rate. During an optimal climate period, the whole society experiences long crop growth periods, high crop production, rapid agricultural development, rapid population growth, and a prosperous economy. However, if the population growth rate is greater than the grain growth rate at this time, a decrease in the per capita grain possession is possible due to an excessive population growth, which can then trigger famine. At this time, the transmission pathway is warm → bumper or normal harvests → prosperous economy/population growth → severe famines. Such a situation may appear during any period from prosperity to decline after a flourishing age. Analogously, economic depressions and social unrest can be caused by climatic pessimum. In the early stages of social restoration, the population base is small; thus, although the grain output is not high, the per capita grain possession is sufficient, which may also reduce the occurrences of famines. At this time, the transmission pathway is cold → poor harvests → depressed economy/social upheavals → population decrease → famine decline. This pathway may appear in the early period of a new dynasty after an alternation of dynasties. As the scholars have revealed about the results of the studies of climate change on social development and human adaptation in the historical period, climate change is the dominant triggers of human tragedies such as massive population decline and unnatural death. Climate change through affected food production subsystem to impact population subsystem and economic subsystem. The results challenge classic Malthusian and post-Malthusian interpretations of historical Chinese social developments and population cycles (Malthus, 1978; Lee et al., 2013, 2014; Pei et al., 2016a, 2016b; Zhang et al., 2011b, 2015).

In the meantime, combined with the social attributes of famines in historical records. Even if there are disasters and a poor harvest, government regulation and relief may prevent famines. In this situation, the transmission pathway is warm → bumper or normal harvests → prosperous economy → more famine relief. Similarly, periods of regime instabilities and social disruptions can be caused by climatic pessimum. The lack of historical records of these periods indicate that the existing historical records have been reduced, resulting in relatively low famine indices. The actual transmission pathway of this situation is cold → poor harvests → depressed economy/social upheavals → less famine records. Therefore, the transmission pathway of climate change → the grain harvest → economy → famine may exist in the transmission pathways dominated by socio-economic factors, but the pathway should be further analyzed in combination with specific historical cases. In addition, the differences in the resolution of the temperature sequence and other socio-economical sequences may affect the accuracy of the final result. This issue as a research focus is the next stage of our research needs attention.

The relationship between climate change and social development of human being is a long-standing research hotspot. From the macroscopic perspective of long period and large space, the scholars tend to discuss the process and mechanism of climate change to social development and human adaption in historical period, for example, economic crisis, peasant uprising, social vicissitude, and human civilization (Pei et al., 2015; Xiao et al., 2015; Yin et al., 2016b; Zhang et al., 2011a; Zheng et al., 2014). Because the historical data have different sources and quantitative methods, there may be differences in research issues, theoretical frameworks, and related conclusions in the study of historical climate change and socio-economic development of different research teams. These differences are not mutually exclusive and contradictory, but mutually complementary.

Conclusion

In this paper, based on a conceptual framework of food security, we selected 10-year resolution sequences of grain harvest grades, famine indices, and economic levels in China over the past 2000 years (206 BC–AD 1911) to quantify and recognize the main transmission pathways of climate change impacts during cold and warm units of climate change. The main results were as follows:

Under the background of climate change, the three social subsystems of the grain harvest, famine degree, and economic level present the characteristics of population increases and economic development during warm periods and population decreases and economic recession during cold periods. Based on the idea of layered transmissions, the transmission relationship can be stated as follows: climate change → grain harvests → famines → economic levels. There are 16 main transmission pathways, with 10 pathways starting from cold units and 6 pathways starting from warm units. In the main pathways for the cold and warm units, the main pathways dominated by natural factors and socio-economic factors, with each factor set accounting for approximately 50%. However, the probability of the main pathways with social influence leading toward negative developments was 60% for cold units, and the probability of the main pathways with social influence leading toward positive developments was 66.7% for warm units.

There were five transmission pathways starting from cold units and three transmission pathways starting from warm units that follow the normal logical reasoning and that were dominated by natural factors. Of all the transmission pathways starting from cold units (100%), 14.7% of the observed depressed economies may have a more direct relationship with climate change. However, 32.3% of the prosperous economies may have a more direct relationship with climate change during warm units (100%).

Some of the main transmission pathways were dominated by socio-economic factors. Of these pathways, for those starting from cold units, there were three transmission pathways that showed economic depressions during bumper or normal grain harvests and mild or moderate famines (i.e. a negative development). There were two transmission pathways that showed severe famines with concurrent bumper grain harvests, while a severe famine did not further trigger economic depression, instead resulting in a normal or prosperous economy (i.e. a positive development). Famine is the most important link in the climate change influence on transmission chains regulated by socio-economic factors. Famines can reflect both the unsafe natural production subsystem caused by poor harvests and unsafe grain consumption caused by the different capacities of grain allocation in the socio-economic subsystem or the elimination of local famines. Thus, there may be another pathway of transmission, such as climate change → agricultural yield/grain production → economic prosperity → famine relief (social vulnerability).

Footnotes

Acknowledgements

We thank the editor and two anonymous reviewers for stimulating discussion and valuable comments on this manuscript. We also thank Yin Jun and Wei Zhudeng for providing the basic data for this manuscript.

Funding

This research was supported by the National Natural Science Foundation of China (Grant no. 41771572).

References

Bai

Kung

(2011) Climate shocks and sino-nomadic conflict. Review of Economics and Statistics 93(3): 970–981.

Buckley

Fletcher

Wang

SYS

et al . (2014) Monsoon extremes and society over the past millennium on mainland Southeast Asia. Quaternary Science Reviews 95(1): 1–19.

Büntgen

Tegel

Nicolussi

et al . (2011) 2500 years of Europe climate variability and human susceptibility. Science 331: 578–582.

Butzer

(2012) Collapse, environment, and society. Proceedings of the National Academy of Sciences of the USA 109(10): 3632–3639.

Chu

(1972) A preliminary study on the climatic fluctuation during the last 5000 years in China. Science China 12(2): 226–256.

DeMenocal

(2001) Cultural responses to climate change during the late Holocene. Science 292: 667–673.

Fang

Yin

et al . (2014a) Methodology for quantifying social and economic series on the base of semantic differential to serve the research on impacts of climate change. Quaternary Science 34(6): 1204–1214.

Fang

Yin

et al . (2015) Transmission of climate change impacts from temperature change to grain harvests, famines and peasant uprisings in the historical China. Science China Earth Sciences 58(8): 1427–1439.

Fang

Zheng

(2014b) Historical climate change impact-response processes under the framework of food security in China. Scientia Geographica Sinica 34(11): 1291–1298.

10.

(2011) Climate Change through Ancient Chinese Dynasty. Beijing, China: Science Press, pp. 126–133.

11.

Fang

Zheng

(2014) Learning from the historical impacts of climatic change in China. Advances in Earth Science 29(1): 23–29.

12.

Zheng

Fang

et al . (2003) Winter half year temperature reconstruction for the middle and lower reaches of the Yellow River and Yangtze River, China, during the past 2000 years. Holocene 13: 933–940.

13.

Haug

Günther

Peterson

et al . (2003) Climate and the collapse of Maya civilization. Science 299: 1731–1735.

14.

IHOPE (2010) Developing an integrated history and future of people on earth (IHOPE): Research plan. IGBP Report no. 59. Available at: http://www.ihdp.unu.edu/docs/Publications/IHOPE/IHOPE_ResearchPlan_2010.pdf.

15.

Lee

(2014) Climate-induced agricultural shrinkage and overpopulation in late imperial china. Climate Research 59(3): 229–242.

16.

Lee

Zhang

(2010) Changes in climate and secular population cycles in China, 1000 CE to 1911. Climate Research 423: 235–246.

17.

Lee

Fok

Zhang

(2008) Climatic change and Chinese population growth dynamics over the last millennium. Climatic Change 88: 131–156.

18.

Lee

Fok

Zhang

(2013) A tale of two population crises in recent Chinese history. Climatic Change 116: 285–308.

19.

Malthus

(1798) An Essay on the Principle of Population. London: Joseph Johnson.

20.

PAGES (2014) PAGES restructured. Past Global Changes Magazine 22(1): 1–56.

21.

Pei

Zhang

(2014) Long-term relationship between climate change and nomadic migration in historical china. Ecology and Society 19(2): Article 68.

22.

Pei

Zhang

Lee

(2016a) Contextualizing human migration in different agro-ecological zones in ancient China. Quaternary International 426(9): 65–74.

23.

Pei

Zhang

et al . (2015) Epidemics in Ming and Qing China: Impacts of changes of climate and economic well-being. Social Science & Medicine 136–137: 73–80.

24.

Pei

Zhang

et al . (2016b) Temperature and precipitation effects on agrarian economy in late imperial China. Environmental Research Letters 11: 64008–64016.

25.

Fang

Yin

(2014) Impact of climate change on fluctuations of grain harvests in China from the Western Han Dynasty to the Five Dynasties (206 BC–960 AD). Science China Earth Sciences 57: 1701–1712.

26.

Liu

Fang

et al . (2016) The relationship between climate change and wars waged between nomadic and farming groups from the Western Han Dynasty to the Tang Dynasty period. Climate of the Past Discussions 12(1): 137–150.

27.

Teng

Fang

(2014) The reconstruction and analysis of famine sequence from the Western Han to the Qing Dynasty (206 BC–1911 AD). Journal of Chinese Historical Geography 29(4): 26–32.

28.

Wei

Fang

(2015a) A preliminary analysis of economic fluctuations and climate changes in China from BC 220 to AD 1910. Regional Environmental Change 15(8): 1773–1785.

29.

Wei

Fang

et al . (2014) A review of climatic impacts on Chinese socio-economic development over the past 2000 years. Advances in Earth Science 29(3): 336–343.

30.

Wei

Rosen

Fang

et al . (2015b) Macro-economic cycles related to climate change in dynastic China. Quaternary Research 83(1): 13–23.

31.

Xiao

Fang

Zheng

et al . (2015) Famine, migration and war: Comparison of climate change impacts and social responses in north China between the late Ming and late Qing dynasties. Holocene 25(6): 191–196.

32.

(2004) The Twenty-Four Histories. Shanghai, China: Chinese Dictionary Press (in Chinese).

33.

Yin

Fang

(2016a) Correlation between climate and grain harvest fluctuations and the dynastic transitions and prosperity in China over the past two millennia. Holocene 26(12): 1914–1923.

34.

Yin

Fang

(2015). Relationships between temperature change and grain harvest fluctuations in China from 210 BC to 1910 AD. Quaternary International 355: 153–163.

35.

Yin

Fang

(2016b) Climate change and social vicissitudes in China over the past two millennia. Quaternary Research 86(2): 133–143.

36.

Yuan

(2008) History of Chinese Disasters. Zhengzhou, China: Zhengzhou University Press (in Chinese).

37.

Zhang

Lee

Wang

et al . (2011a) Climate change and large-scale human population collapses in the pre-industrial era. Global Ecology and Biogeography 20(4): 520–531.

38.

Zhang

Brecke

Lee

et al . (2007) Global climate change, war, and population decline in recent human history. Proceedings of the National Academy of Sciences of the USA 104(49): 19214–19219.

39.

Zhang

Lee

Wang

et al . (2011b) The causality analysis of climate change and large-scale human crisis. Proceedings of the National Academy of Sciences of the USA 108(42): 17296.

40.

Zhang

Pei

Lee

et al . (2015) The pulse of imperial china: A quantitative analysis of long-term geopolitical and climatic cycles. Global Ecology & Biogeography 24(1): 87–96.

41.

Zhao

(1977) The Draft of the Qing History. Beijing, China: Zhonghua Book Company (in Chinese).

42.

Zheng

Xiao

Fang

et al . (2014) How climate change impacted the collapse of the Ming dynasty. Climatic Change 127(2): 169–182.