Between 1876 and 1878 a large-scale drought occurred in China. This is a major meteorological disaster and an extreme climate event despite the cold climate at the end of the Little Ice Age. In this paper the dynamic evolution of the occurrence and development of the drought is reproduced on the basis of historical literature records. These were used to calculate the yearly numbers of drought-hit counties and to determine the spatial distribution in addition with concomitant famine, locust plague and pestilence epidemic for each of the three years. The persistent drought disaster spread over 13 provinces with its center in Shaanxi, Henan and Shanxi provinces, where the continuous non-soaking rain period exceeded 340 days. Conclusively, it is more severe than the worst drought (1928–1930) in the 20th century. This drought disaster of 1876–1878 took place in the descending phase of the 11th sunspot activity period and the start of the 12th period. It also happened during a spell of frequent El Niño events and corresponds with an extremely strong El Niño.
extreme climate event ; drought ; meteorological disaster ; 1876–1878
Over the last 1,000 years 15 major droughts have been observed [ Zhang, 2004 ; 2005a ] occurring in different cold and warm climate phases [ Zhang , 2005b ]. The 1784–1786 dry event emerged in a relatively warm climate phase [ Zhang , 2000 ]. This drought happened in the 2nd to 4th reigning year of the Guangxu Emperor of the Qing Dynasty. It is a long-lasting extensive disaster, bringing about extremely serious social consequences and is thus called the Guangxu Severe Drought by Chinese historians. The historical records subjected to rigorous emendation are employed to analyze the 1876–1878 drought in a dynamic manner, including its genesis and development, severity, resulted damages and losses, and the related climate characteristics in order to enhance the understanding of extreme climate aridity in China.
The 1876–1878 heavy drought event hit parts of 13 provinces, i.e., Liaoning, Inner Mongolia, Hebei, Shanxi, Shaanxi, Henan, Shandong, Gansu, Ningxia, Sichuan, Hubei, Anhui and Jiangsu. It started in January 1876 and ended in mid August 1878 (all months refer to the months in the lunar calendar). The authors utilize the historical climate descriptions quoted from Compendium of Chinese Meteorological Records of the Last 3,000 Years [Zhang, 2004b ] to retrieve the processes of occurrence and development of the hazard, along with the calculation of dryness characteristics.
In 1876, a summer drought was experienced in northern China. Hebei, Shanxi, Henan, and Shandong in conjunction with parts of Liaoning and Inner Mongolia suffered little or no rainfall starting from January 1876 to the start of the wet season in mid July 1877. Then the drought was relieved by some degree from south to north. Especially Cangzhou and Gaocheng in central Hebei province experienced the drought for almost one year (starting from autumn 1876 and ending on July 8, 1877). The hazard persisted much longer in Shandong province, with Linju, Shouguang and Shanxian counties suffering until late August 1877. The summer dryness of 1876 prevailed over Jiangsu and Anhui provinces, over the Jiang-Huai Valleys, and the lower reaches of the Yangtze River. The summer drought continued at Yingshan of Hubei and Nanchong of Sichuan in the mid-upper basin of the Yangtze River until autumn 1877. Summer aridity was also experienced in Gansu province in 1876.
In 1877 the drought-stricken area was expanded and the hazard lasted even longer, leading to severe aridity and no rainfall in summer and autumn over Hebei and Shanxi, with the longest rain-free period in history. In Pinglu of Shanxi, for example, “No rains occur for more than 200 days from spring to winter” [ Zhang, 2004b ]. In Hengshan of Shaanxi province and its vicinity for instance, the dryness was severe enough to have “Not even a single raindrop from 1877 spring to summer” [ Zhang, 2004b ] and the drought lasting until the spring of 1878. The summer to autumn drought of southern Shaanxi in 1877 continued till 1878. Many counties of Gansu and Ningxia reported that “Not a single raindrop was received throughout the year”, or “No rainfall in May until the end of the year occurred”. As shown in the local chronicles, Kaifeng, Fangcheng, and Mianchi of Henan province experienced “Severe successive droughts in summer, autumn and winter of 1877 and then persisting till the end of 1878 spring”. Spring dryness was severe all over Shandong province. In addition, summer drought pounded parts of the provinces of Anhui, Jiangsu and Hubei. The disaster was also severe in northern Hunan, especially in Cili and Shimen counties where “Very serious aridity lasts for 70 days”. The dry situation of Sichuan, particularly in the northern part of the province, was also in abundance following the report of “No rain from April to August” in 1877. In Guizhou province the situation was similar compared with Sichuan, where they noted that “No rain is received from October 1876 to April 1877”.
In 1878 the affected areas appeared to decrease in Hebei, Shanxi, Inner Mongolia, Henan, and Shandong provinces, with a westward movement of the main drought-stricken areas towards Shaanxi, Gansu and Sichuan. There was little or no rainfall in spring and summer over Hebei, Shanxi, Henan and Shandong, with the drought coming to an end when precipitation started in mid August 1878. Later on, the rain band remained in Shanxi, Henan and Shandong from the end of September to mid October 1878, bringing about heavy rainfall. Until then the nearly three year long aridity was thoroughly overcome while the following more than decade-long precipitation caused field crops to mould and rot in Yuci, Xiyang and Shouyang of Shanxi province.
Documentary evidence from Zhang [2004b] indicates that for central Shaanxi the drought, starting in April 1877 and persisting one year, was fully relieved due to multi-day heavy rainfall beginning in 13 April 1878. But the drought subsequently reoccurred in May to June and was so severe that the “Runde Spring of Qishan becomes dried up once more”. The local chronicles showed that Wudu, Tianshui, and Yongdeng of Gansu suffer grave dryness in autumn and winter of 1878 which last until the early summer of 1879. Besides, another summer to autumn drought happened in Sichuan, with particularly Ya’an suffering “Serious aridity in May which persists for the following 70 days”. Such disasters affected other areas as well. The long lasting extensive dryness terminated in the whole country in 1878.
Historical drought events (with the older county names converted into the present-day ones) are employed to prepare maps to depict the dynamic development of drought-hit areas on annual basis. Figure 1 presents the drought-hit areas of the 1876–1878 drought (brown shading) and it shows the largest drought-hit area in 1877, especially located in Shanxi province.
Drought-hit areas in 1876 (a), in 1877 (b), and in 1878 (c) in China (the blue contours denote the number of days without crop-required rainfall)
Aridity can be quantitatively denoted by the number of days without rainfall, the number of continuous no-rain days, and departure of precipitation. And the aridity index is sometimes constructed by means of meteorological data. These methods are, however, unsuitable for the use in calculating the index from historical descriptions. In addition, the concept of a period of days of non-soaking rainfall is employed to represent the field water deficit and hence aridity, which is used in various kinds of reports as concerning aridity. In historical documents and books there are descriptions of severe local droughts, as given by the phrase of “Without any raindrop from a particular month to a later month”. However, “non-soaking rain” (NSR) descriptions denote only the day or a spell of days having insufficient rainfall for crops, differing from the definition of rain-free day or a spell of rainless days in modern meteorology. In addition, the described NSR varied spatially in amount, depending on the season and growth stage of crops. As a result, it is impossible to get the actual rainwater amount from the NSR days in a period for a particular region. Despite this, the NSR connotations are comparable in historical and modern times. The authors utilize the concept of a spell of NSR days used in historical climate research to represent the aridity intensity.
Based on the records of “no rain date” and “date of rain starting”, the length of NSR days can be calculated. In Pinglu of Shanxi province, for instance, the local chronicle noted that “No precipitation during more than 200 days from spring to winter occurred”. Based on that, we determined 200 NSR days for this county. Likewise, based on the description of “No rain from June to the following May” for a particular place, we can infer that there are more than 330 NSR days in total for drought persistence, and of which 160 days were in the previous year. The calculated lengths of regional NSR days are presented in Table 1 . It is shown that such days exceeded 200 over the mid basin of the Yellow River and southern Shaanxi in 1877, and the duration of NSR days was more than 300 in 1877 to 1878, compared with 340 at Huayin of Shaanxi.
|County||Length (1877)||Length (1877–1878)||Historical records (in the lunar calendar)|
|SA Pinglu||> 200 days||Rain-free length of more than 200 days from spring to winter|
|SA Gaoping||≈ 160 days||> 330 days||No rain from June to following May 1878|
|SA Fenxi||> 200 days||No rain from spring to September|
|SA Linyi||> 250 days||No precipitation after March|
|HN Kaifeng||170 days||285 days||No rain from June 1877 to following 14 March 1878|
|HN Fangcheng||> 200 days||> 300 days||No rain from May to following March 1878|
|HN Mianchi||240 days||> 300 days||No rain from spring to following March 1878|
|SAA Huayin||250 days||340 days||No rain from April to following 11 March 1878 and on the day heavy rain occurs|
|SAA Hengshan||180 days||Not a single raindrop falling from spring to summer|
|SAA Fugu||180 days||No rain from spring to autumn|
|GA Tianshui||> 200 days||No rain from May to the end of the year|
Comparison is made with the continued NSR records of the 1876–1878 drought to the 1928–1930 most severe drought, as the most disastrous event in the 20th century in China. The 1928–1930 drought is determined in terms of the NSR descriptions including “Not a single raindrop from spring to autumn” in 1928 in Shaanxi, “No rain from spring to autumn in 1928 in southern Shanxi” [ Wen , 2008 ], “From spring to late summer in Gansu” in 1928, and “No rain for months from spring to summer at Luoyang” in the 1929 local chronicle [ Wen , 2008 ]. The 1928–1930 drought was not so long as the 1876–1878 counterpart at Huayin of Shaanxi and was shorter compared with the re-cords of more than 200 rain-free days from spring to winter at Pinglu and Gaoping of Shanxi, as shown in Table 1 . Just from the mentioned NSR lengths it is obvious that the 1876–1878 aridity exceeded its equivalent in 1928–1930.
The occurrence of dried-up rivers, lakes, wells and springs in 1876–1878 is also well described in historical records in northern China. Because the hydrological regime of modern times is affected by water conservancy works and sharp increases in water-use for agricultural purposes, it is impossible to contrast the historical descriptions with records of water levels in rivers and lakes of modern times. However, the past descriptions remain interesting, e.g., “The Fenhe River was dried up and the Hanshui River permitted people to cross on foot”, phenomena which did not occur in modern times. The descriptions of such dry-ups are listed in Table 2 .
|1876||SD Shouguang||Mishui R.||Heavy drought in spring; Mishui is dried up|
|1876||SD Laiwu||Wenhe R.||The Wenhe River has run dry|
|1877||SA Jiangzhou||Kuaishui R.||The river runs dry both in June and July, each lasting more than 10 days|
|1877||SA Quwo||Fenshui R., Kuaishui R.||Both rivers run dry in June|
|1877||SAA Huaxian||Baiya L.||Heavy drought and the Baiya Lake is dried up|
|1877||SAA Yangxian||Most wells run dry|
|1877||SI Hechuan||Qujiang R.||No water left in the Qujiang River|
|1878||HB Laohekou||Hanshui R.||No water left in the river in summer|
|1878||HB Jingshan||Hanshui R.||Severe drought, crossing the Hanshui on foot is possible|
|1878||SAA Qishan||Runde S.||Drying-up once more|
Note: Data source [ Zhang, 2004b ]
The 1876–1878 drought had crucial impacts on crops which resulted in poor and even no yields. Hence, the drought was causing serious famine, accompanied by spreading pestilence and locust plagues, leading to considerable damages to national resources and great sufferings of the people.
During the long lasting and extensive drought the provinces of Shanxi and Henan experienced the most adverse impacts. As written in the Encyclical of the Emperor Guangxu in August 1877, “Shanxi is suffering a lot from the drought and compared to it, Henan is suffering less, and of the 400,000 Liang (1 Liang=50 g) relief silver, 70% is allotted for Shanxi, the rest is given to Henan”. The report on 8 November 1877 submitted by Shanxi officials stated that “76 counties of Shanxi province have experienced the severe drought. The field crops are being withered because of water deficit for a long time and we have ordered peasants to replant buckwheat and other course food crops as a remedy. Unfortunately, from summer to autumn we are receiving frequently reports from the inferiors, claiming that strong winds blew about dark clouds or only very light rain fell onto ground, infiltrating soil, one or two Cun (1 Cun=3.3 cm) in depth only. It is scorching and we have the feeling over fire. The newly planted crops like buckwheat and other coarse suffer withering when coming up out of surface, and peasants are losing their hope of harvest” [ Zhang, 2004b ].
Due to the 1876 drought grain prices were high, hence, resulting in a fast spread of famine throughout the provinces of Hebei, Shanxi, Shaanxi, Inner Mongolia, Liaoning and Shandong, along with parts of Jiangsu, Anhui and Henan. In 1877 the crop production failed and famine hit an even bigger area, leading to a situation in which “The dead are seen lying along roads and passengers are trying to kill others for sustenance”. The picture was miserable in Jiangxian county of Shanxi, where “Victims knife human corpses, dig out dead bodies for sustenance, even have meals exchanged corpses of their own children and even kill live persons as food for survival so that 40%–50% of the population has died of starvation”, and in Jishan county “There are villages with all people or nearly all killed by the demon of drought”. The situation of Henan was even more tragic, where “An expanse of no green plants stretches for thousands of Chinese Li (1 Li=500 m), a large number of the dead are lying in ditches and valleys, the rest surviving on relief”, and “No people surviving in nine of the ten families”. Again, “In western Henan province including Shanzhou, the belts along the Yellow River and Ruzhou prefecture more than half the population died out of starvation, and more than a million people surviving famine are now begging in Xinyang and its vicinity in southern Henan”. In Jinbian of Shaanxi, “With no grass and bark left, people are surviving on a kind of white earth, leading to the fact that the old and young have lost their lives out of too much soil in stomach and only the strong adults survive, and some of the victims living even on killed humans”. Such terrible records exceed 80 entries in the historical literature, with the related counties marked by blue dots in Figure 2 . In Huayin of Shaanxi, the description runs “Countless people have starved to death. When autumn crops are harvested, almost half the population died out of gulping. Since then more than 50 years has elapsed but the population has not reached the pre-drought level” (Fig. 2 a). In the spring of 1878 the famine was in its strongest phase, when “Two thirds of the dwellers died from hunger in summer at Pucheng county of Shaanxi”. The famine came to an end (Fig. 2 b) after the harvest of autumn crops in the five provinces of North China in 1878.
Distribution of famine (pink), locust plague (brown), and pestilence (red) areas, and location of the sites where victims survived on human corpses (blue dots) in 1877 (a) and 1878 (b)
With ever-increasing famine, pestilence took place in 1877 (Fig. 2 ), with the northern epidemic areas dominantly in Hebei, Shanxi, and Henan, and, to less degree, in Shandong and Liaoning. The epidemic regions expanded in 1878, when the number of affected prefectures and counties more than doubled. “The pestilence is so strong that patients are prone to death” as written in the local chronicle of Jiangxian, Shanxi province, while for Jishan of the same province “Pestilence is spreading from summer to autumn, with the corpses piling up”. And for Xinyang of Henan the description is “Pestilence breaks out so terribly that the people are dying one after another in spring 1878. The survivals from the disaster are likely infected again. The relief gentlemen are busy with their affairs. Death occurs occasionally and becomes less and less after June”. The pestilence plague was eradicated only when the vast expanse drought and famine ended in 1879.
This locust plague caused terrible havoc in 1876–1878, especially in 1877. The plague-stricken band covered Gansu, Hebei, Henan, Shandong, Anhui, Jiangsu, Zhejiang, and Hubei (Fig. 2 ). The insects lived on crops, which were almost vanished, leading to an even more serious famine. Just as described in the local chronicle of Xinyang county of Henan as “In this year locust plague broke out, first living on crops over the basins of the Yellow River and the Luoshui River and devastating a vast expanse of growing crops. They multiply quickly and fly southward in swarms downwind when no crops are available”. There is even such a description, stating that “The swarms passed through Xinyang for successive three days and nights, with the largest group covering tens of Chinese Li in the horizontal dimensions to produce a dark sky”. There is a story saying that “In the band surrounding the Bohai Sea, consisting of Tianjin, Wuqing, and Jinghai, locusts covered the ground in June and even crossed rivers in a rolling ball as a marvelous spectacle” [ Zhang, 2004b ].
The severe drought event of 1876–1878 occurred when the cold climate of the Little Ice Age was about to end and the Northern Hemisphere climate was about to experience quick warming [ IPCC, 2001 ], during which period the climate over parts of Europe and North America was beginning to become warm but East Asia remained in cold [ IPCC , 2007 ].
During 1876–1878 the precipitation pattern was marked by dryness (wetness) in the north (south), with drought-stricken areas being extensive and reaching the middle and lower Yangtze River Valley in the south. The rainy seasons of these 3 years were abnormal in eastern China. Cold air masses were highly active and few typhoons occurred in the littorals.
The rainy seasons of 1876–1878 began later and brought little precipitation to eastern China. From historical records [ Zhang , 2004b ] we know that the rainy season started on 19 July 1876 in the Huanghuai Area, which is later than the mean of 1961–1990 (all of the mean values of 1961–1990 used in this paper are from CMA surface observational datasets). The 1877 description was “No rain from April to July” for the mid-lower reaches of the Yangtze River, indicating that the rain was of extreme anomaly, i.e., the rainy season (Meiyu) did not occur. In 1878 the rainy season began on 12 August in North China, the date being much later compared with the mean of 1961–1990, but the total precipitation was enough to end the drought.
In 1876 the first frost took place on 27 September, covering a vast expanse of Shanxi, Hebei and Shandong provinces, with its southern boundary in Xixian, Jingxian and Ningjin counties. In comparison with the mean value of 1961–1990, the early frost day was 16 days ahead① .
The frosts in 1877 in central Shanxi province showed severe impacts because the first frost event was experienced much earlier compared with 1961–1990. For example, at Heshun of Shanxi, an “Early severe frost killing field crops” happened as early as 21 August 1877, 31 days ahead of the average of first frost days of 1961–1990 (22 September) and 19 days ahead of the earliest frost record (8 September).
The 1876 cold air masses came earlier with high severity, leading to intense snowfall over Shandong in early November. Vigorous cold waves were frequent in the 1877/1878 winter.
Accompanied by more than 60 days of snowy and icy weather, starting at the end of December, the hazard spread throughout North China, into provinces of the mid-lower Yangtze River Basin and even to South China. The historical descriptions are as follows. The counties of Pinglu of Shanxi and Zhucheng of Shandong, both close to 35°N, suffered “Extremely cold and icy weather to cause wells to be frozen”. The Penglai Isle “Has been cut off from the land for as many as three months”. The Hanshui River in Hubei province and the Dongting Lake in Hunan were frozen, as recorded in the Xiangyin chronicle. There is a description of “Sudden strong cold freezing tree to death” and “Frigid winter causing rivers to be frozen to the bottom” in Jiangsu province. Strong cold waves attacked one after another the zones south of the Nanling Mountains (south of 25°N), leaving more than one description that “Bleak winter lasts the month, fish is frozen to death and the cold ends only after two months”. The event of an intense cold wave on 3–4 January, 1878 affected Shaoguan in Guangdong with “Heavy snowfall and icy weather for a period of days to cause the death of cattle, sheep and goats”, and in the Pearl River Delta “Snowfall and frost come together to induce freezing weather for two months”. Hazardous graupel, snow and freezing extended south into the Leizhou Peninsula. The conditions of the harsh winter of 1877/1878 were seldom seen in the records of 1961–1990. The cold air mass activity in autumn and winter of 1878 began early, with snowfall occurring in Gaoping of Shanxi province on 4 October, which was over one month earlier than the earliest date (2 November) and the mean date (16 November) of first snowfall day of 1961–1990.
The 1876–1878 drought was related to the external factors and the corresponding relations are the same as those reported in previous studies [ Zhang, 2000 , Waldmeier, 1961 , Wang and Zhao, 1987 , Chen, 1991 , Zhang and Xue, 1994 , Zhao, 1996 and Zhang et al., 1999 ].
The studied drought event occurred between the 11th and 12th period of solar activity, i.e., between the decreasing phase of sunspot’s number in the 11th period and the year of minimal number in the 12th period, with the annual mean of 3.4 [ Waldmeier, 1961 ]. The period of the 1876–1878 disaster were, respectively, in relation to the phases of m –2 and m –1, and extremely small number of sunspots. The years 1876–1877 correspond to less rainfall in China, a conclusion that is in agreement with that given by Wang and Zhao  . 1878 was a year of sunspot extreme, with torrential rains after mid August over North China that relieved its aridity, followed by another period of heavy rain, causing a plethora of precipitation, a result that is also in harmony with the conclusion of Wang and Zhao  .
No detailed observations of pressure and SST were made when the 1876–1878 drought happened in China, but the chronological table of the past El Niño events [ Quinn and Neal, 1992 ] will be used. The severe dryness happened in the period with frequent El Niño events, with an El Niño of S++ level (i.e., very strong) emerging in 1877–1878. 1879 was the first non-El Niño year after the preceding El Niño event. The years 1871 and 1874 were years of El Niño events at S + level (strong) and M level (moderate-intensity), respectively. The drought period was followed by another M level El Niño event in 1880. It can be seen that these episodes were at an interval of 2–3 years, a result which is in accordance with those studies [ Chen, 1991 , Zhang and Xue, 1994 , Zhao, 1996 and Zhang et al., 1999 ; Zhang and Akimasa, 1994]. A linkage of El Niño events to precipitation in China can be drawn.
According to the records of past volcanic activities in the world [ Simkin and Siebert, 1994 ], there were numerous vigorous eruptions at high latitudes before 1876–1878. In Iceland, for instance, the Askja volcano erupted in March 1875 with a VEI2② of 5 (very large), the Grimsvotn erupted in January 1873 with a VEI of 4 (large), and the Sinarka volcano in the Kurile Islands erupted in 1872 with a VEI of 4. Besides, volcanic activity occurred also at mid latitudes, the example being the 1877 Suwanose-Jims volcanic eruption (29.5°N, 129.7°E) with a VEI of 4 in Japan. Whether the eruption materials had influences on the atmospheric circulation over East Asia and hence on the extreme drought event in China is unknown.
The large-scale severe drought of 1876–1878 in China can be regarded as a major climatic disaster and an extreme climate event occurring when the Little Ice Age was about to end (at that time China remained in a cold climate) and large-scale warming in parts of the Northern Hemisphere (Europe and North America) began [ IPCC, 2001 and IPCC, 2007 ]. This event hit a total of 13 provinces of China, with the main affected areas in Shanxi, Henan and Shaanxi, where the duration of NSR was above 200 days in 1877, and of above 300 days for the years 1877–1878. Regarding the length of NSR, the studied event exceeded the longest drought event (1928–1930) observed in the 20th century.
The 1876–1878 drought event was accompanied by or resulted in famine, locust plague and pestilent epidemic, which gave rise to very serious social consequences and was thus a typical case of climate disaster in relation to social development.
The results of the analysis of the drought disaster in association with the climate conditions showed that the drought happened in a low-value phase of solar activities. The drought can also be associated with El Niño events, as the severe dryness occurred simultaneously with an extremely strong El Niño event. In addition, vigorous volcanic eruptions occurred ahead of the 1876–1878 disaster in high latitudes. In comparison with the aridity event of 1784–1786 taking place in a relatively warm climate phase [ Zhang , 2000 ], each of the disasters was associated with an El Niño episode and major volcanic activities, but at different phases of solar activity.
This work is supported by the National Science and Technology Support Program of China under Grant No. 2007BAC29B-01, and Grant No. 2010CB950103 under China Global Change Research Program.
①. The mean first frost day is 13, 27 and 15 October and the earliest first frost day is 29 September, 9 and 2 October, respectively, for Xixian, Jingxian and Ningjin counties and the data of first frost & snowfall come from CMA surface observational datasets.
②. VEI (Volcanic Explosivity Index), a simple 0–8 index of increasing explosivity. VEL=5 means its general description as very large, and VEL=4 means large.