Rain Belt and Flood Peak: A Study of the Extreme Precipitation Event in the Yangtze River Basin in 1849

Abstract

Floods caused by extreme precipitation events, in the context of climate warming, are one of the most serious natural disasters in monsoon region societies. The great flood in the Yangtze River Basin in 1849, in Eastern China, was a typical extreme flood event. According to historical archives, local chronicles, diaries, and historical hydrological survey data, this study reconstructed the temporal and spatial patterns of extreme precipitation in 1849, and the flood process of the Yangtze River. We found four major precipitation events at the middle and lower reaches of the Yangtze River, from 18 May to 18 July 1849. The torrential rainfall area showed a dumbbell-like structure along the Yangtze River, with two centers distributed separately in the east and west. For the specific flood process of the Yangtze River, many tributaries of the Yangtze River system entered the flood season consecutively since April, and the mainstream of the Yangtze River experienced tremendous pressure on flood prevention with the arrival of multiple rounds of heavy rainfall. In mid-to-late July, the water level and flow rate of many stations along the mainstream and tributaries had reached their record high. The record-breaking peak flow rate at many stations along the mainstream and tributaries in the middle reaches of the Yangtze River indicated intense precipitation in the area. The heavy rainfall disaster in the Yangtze River Basin could be driven by these reasons. First, the cold air in North China was extraordinary active in 1849, which made it difficult for the subtropical high pressure to move northward. Second, the rain belt stagnated in the Yangtze River Basin for a long time, and the Meiyu period reached 42 days, 62% longer than normal years. Third, the onset of a southwest monsoon was earlier and more active, which provided abundant moisture to the Yangtze River Basin. The great flood disaster was caused by heavy precipitation at the middle reaches, which made it quite different from the other three great floods in the Yangtze River in the 20th century. At present, the large water conservancy projects in the Yangtze River are mainly designed for flood problems caused by rainstorms in the upper reaches of the Yangtze River. The middle reaches of the Yangtze River, however, are facing the weakening of flood diversion capacity, caused by social and economic development. Therefore, future flood prevention measures in the Yangtze River should pay great attention to the threat of this flood pattern.

1. Introduction

Floods are one of the most destructive and influential natural disasters, which can threaten the natural environment and human societies. In the past few decades, the proportion of flood disasters in global natural disasters has been rising [1]. A flood disaster is a comprehensive response of many driving processes, including meteorological forcing, basin morphology (altitude, slope), geological characteristics, and human activities [2,3,4].

Extreme precipitation is the main cause of flood disasters [5]. Studies in Europe, North America, South America, and the Mediterranean have demonstrated that extreme precipitation is the result of multi-scale weather systems (including large-scale atmospheric circulation, mesoscale convective systems, and local convective weather), among which the local convective weather triggered by the large-scale and mesoscale systems is the most important one [2,6,7,8,9].

The latest IPCC research report indicated that the frequency and intensity of heavy precipitation events have increased since the 1950s, for most land area [10], including Europe, North America, Australia, Japan, China, and other regions [2,11,12,13,14]. The increase in the frequency and intensity of heavy precipitation has caused great impacts on river hydrology, and river flood events are occurring more frequently [15,16,17]. On the one hand, human activities could strengthen the river flood control system and improve the flood control capabilities in certain aspects. On the other hand, human activities might also weaken the flood control capabilities in certain areas [18,19,20]. The impacts of increased uncertainty in extreme precipitation and human activities have led to certain types of extreme precipitation patterns that can cause great pressure on river hydrology, thereby increasing the risk of regional flood control. However, the current data and experience are insufficient to identify and respond to such events.

Since the precipitation data available in most parts of the world cover less than a hundred years, it is essential to rely on other proxy data in research. To promote better understanding of the pattern of extreme precipitation and its long-term social impact, it is necessary to conduct more research using other proxy data. Until lately, many scholars have attempted to reconstruct extreme precipitation events in historical periods. Based on comprehensive historical data, Herget et al. indicated that extreme rainfall mainly occurred in the mainstream of the Rhine River, and the Danube, Elbe, Weser and Lahn rivers in Central Europe, in July 1342 [21]. Elleder utilized non-instrumental data to reconstruct the flood process of extreme precipitation in the Czech Republic in 1784 [22]. McEwen and Werritty relied on memorials to reconstruct the hydrometeorology cause of the flood in Britain in 1829 [23]. Based on instrumental and non-instrumental records, Brázdil et al. reported that the heavy precipitation and snowmelt caused by the eruption of the Lakagígar volcano in Iceland in 1783 resulted in three stages of extreme flooding in Britain, France, Hungary, and other countries in 1783 and 1784 [24]. Similarly, Roessler et al. found that the eruption of Tambora led to the abundant snow storage in Europe in 1816 and 1817, and the excessive snowmelt triggered severe floods in Switzerland in 1817 [25]. These studies have enriched our understanding of extreme precipitation events in historical periods.

The Yangtze River is the largest river in China, which originates from Geladandong (6500 m above sea level) in Tanggula Mountain on the Qinghai–Tibet Plateau, and flows from west to east into the East China Sea, through 11 provinces in China (90°33′–112°25′ E and 24°30′–35°45′ N) [26] (p. 1). Its main stream has a total length of more than 6300 km, ranking third in the world, and its drainage area is 1.8 × 10⁶ km² [26] (p. 1). The upper, middle and lower reaches of the Yangtze River are bounded by Yichang in Hubei (62 m above sea level) and Hukou in Jiangxi (25 m above sea level), and the main tributaries of the middle and lower reaches of the Yangtze River include the Han River, the Yuanshui River in the Dongting Lake water system, and the Xiangshui River and the Gan River in the Poyang Lake water system, which are shown in Figure 1. The topography of the Yangtze River Basin decreases from west to east. There are many high mountains and valleys in the upper reaches, with an average gradient of 0.61‰. The middle and lower reaches are mainly distributed with flat alluvial plains, with average gradients of 0.03‰ and 0.007‰, respectively [26] (p. 2). Although complete dike systems are built along the river banks (including main dikes, branch dikes, and beach dikes, etc.), the middle and lower reaches face the most serious flood threat. The Yangtze River Basin is mostly located in the subtropical monsoon zone. With a current annual average rainfall of 1100 mm (of which 70–90% concentrates in May–October), the annual average flow of the river is about 9.6 × 10¹¹ m³ [26] (p. 6). Typically, the flood process lasts for 10 days for a major tributary of the Yangtze River. For the main stream, in which all tributary floods converge, the flood process lasts even longer, usually 20–30 days in the upper reaches, and more than 50 days in the middle and lower reaches [26] (p. 6). The middle and lower reaches of the Yangtze River are important agricultural areas and grain-producing areas, with fertile yellow cinnamon soil and red soil [27]. The area of the middle and lower reaches of the Yangtze River accounts for 43.7% of the total basin area, which is home to more than 400 million people [28] and contributes approximately 73.5% of the gross domestic product (GDP) of the whole basin [26] (p. 10). Moreover, the area of the middle and lower reaches of the Yangtze River is also one of the areas with the highest level of urbanization, the most developed industries, and a key transportation hub of the entire country of China [26] (pp. 19–21).

Since the beginning of the 21st century, three “100-year event” extreme floods (1931, 1954, 1998) have occurred in the Yangtze River Basin, and the latter two resulted in several record highs of water levels and peak discharges at some hydrological stations along the river [29]. In 1954, the abnormal atmospheric circulation in the Northern Hemisphere caused a long-term southerly distribution of the subtropical high, which led to sustained heavy precipitation in the Yangtze River Basin [30], resulting in a high flow rate of 66,800 m³/s at Yichang station, whose annual average flow was only 14,300 m³/s [31]. Driven by the abnormal southward retreat of subtropical high pressure, caused by the abnormal atmospheric circulation and the strongest El Niño phenomenon in the 20th century in 1997–1998, the “twice Meiyu” occurred in the Yangtze River Basin and caused extreme floods in 1998 [32,33,34]. In the same year, the flood peak flow rate of Yichang station, Shashi station, and Zhicheng station reached 63,600 m³/s, 53,700 m³/s, and 68,000 m³/s, respectively [35]. The direct economic losses caused by the extreme floods amounted to 30 × 10⁹ dollars [29], accounting for 3% of the GDP of China in 1998.

The great flood in the Yangtze River Basin in 1849 was a typical extreme flood event. Many studies have investigated this event [36,37,38,39,40], and have reached some consensus on the spatial distribution of flood disasters and other issues. However, there are still some controversies on the temporal and spatial processes and weather conditions of this event, and a reconstruction of the flood process of the Yangtze River is also needed, which can help to understand the specific process and causes of flood disasters. Motivated by these shortfalls, we reconstructed the precipitation process, the center of heavy precipitation, and the flood peak process of the Yangtze River’s mainstream in 1849. The reconstruction was performed by gathering several newly discovered weather diaries and flood survey data that were not used in previous studies, as well as further excavation of archive data from 1849. Based on our reconstruction work, we further analyze the meteorological conditions and disaster types of the 1849 flood, in order to gain a deeper understanding of the taxology of extreme flood events in historical periods and provide practical guides for flood risk reduction in the Yangtze River Basin.

2. Materials and Methods

2.1. Materials

The records used in this study are listed in

Table 1. Types and characteristics of materials utilized in this study.

Type of Materials	Data Name	Author/Publisher/ Collector	Characteristics of Materials
			Time	Space
Archives	Records of Floods of the Yangtze River Basin and the International Rivers in Southwest in the Qing Dynasty [41]	Department of Water Resources Management and Department of Science and Technology of Ministry of Water Resources and Hydroelectric Power of the PRC	High resolution, mainly in days, ten days and months	High resolution, primarily provinces and counties
	The Transcript of Memorials from Ministry of Defense [42]	China’s First Historical Archives
	Collected Grand Edicts in the Daoguang Reign of Qing Dynasty [43]	China’s First Historical Archives
Local chronicles	A Compendium of Chinese Meteorological Records of the Last 3000 Years [44]	Zhang De’er	Low resolution, mainly based on the year, and “record the extremes but not the normal”,	High resolution, mostly provinces and counties
	The Records of Yangtze River:river systsystem [26]	Ministry of Water Resources
	The Records of Yangtze River: flood control [45]	Ministry of Water Resources
	The Water Conservancy Chronicle of Nanjing City [46]	Nanjing Local Chronicles Compilation Committee
	The Water Conservancy Chronicle of Jiujiang City [47]	The Water Conservancy Chronicle Compilation Committee of Jiujiang City
	The Water Conservancy Chronicle of Anqing City [48]	The Water Resources Bureau of Anqing City
Diaries	Guan Tingfen’s Diary [51]	Guan Tingfen	Very high resolution, mostly based on the day and hour, but there are cases of omissions	Very high resolution, mainly counties
	Ping Lu’s Diary [52]	/
	Sanyuantang’s Diary [53]	Zhao Yanchen
	Flood Diary in Ji You [54]	Yao Ji
Survey data	The Compilation of Survey Data of Historical Great Floods in China [55]	Luo Chengzheng	Record extreme precipitation events	High resolution
	The Great Flood of 1954 in Yangtze River [56]	Hydrological Bureau of Yangtze River Water Resources Commission

, which mainly include archives, local chronicles, diaries and survey data.

The archive “Records of Floods of the Yangtze River Basin and the International Rivers in Southwest in the Qing Dynasty” [41] systematically collects rainfall and disaster information from the court archives reported by local officials in the Qing Dynasty. “The Transcript of Memorials from Ministry of Defense” [42] is a transcript of the memorials that preserve the rainfall and disaster information reported by local officials in the Qing Dynasty, which is currently preserved in China’s First Historical Archives. “Collected Grand Edicts in the Daoguang Reign of Qing Dynasty” [43] is a collection of orders issued by Emperor Daoguang of the Qing Dynasty who reigned from 1820 to 1850, which systematically records the disaster and relief measures of each county. Based on these historical records, this study reconstructed the temporal and spatial processes of rainfall and disaster in various counties.

The local chronicle “A Compendium of Chinese Meteorological Records of the Last 3000 Years” [44] systematically collects the meteorological disasters recorded in more than 8000 existing local chronicles in China, which are local records from independent sources that are different from archives. There are 220 records related to precipitation in the middle and lower reaches of the Yangtze River in 1849, which can supplement the understanding of rainfall and disasters in various counties. Furthermore, this study also utilized the flood and hydrological data of various counties in 1849 from conservancy chronicles, such as “The Records of Yangtze River: river system” [26], “The Records of Yangtze River: flood control” [45], “The Water Conservancy Chronicle of Nanjing City” [46], “The Water Conservancy Chronicle of Jiujiang City” [47] and “The Water Conservancy Chronicle of Anqing City” [48].

The diaries with clear locations, dates and weather descriptions are important for high-resolution climate reconstruction research [49,50]. We gathered four diaries [51,52,53,54] with continuous and detailed weather records in 1849. According to these diaries, we systematically extracted daily precipitation and disaster records, which provided the basis for determining the Meiyu (Meiyu is a unique weather phenomenon) period in 1849.

The survey “The Compilation of Survey Data of Historical Great Floods in China” [55] is a national survey of flood historical data in river reaches by the Chinese Water Conservancy Department to supplement hydrological data from the 1950s to 1970s. It contains flood survey information of 5544 river reaches, including the historical flood peak data of these reaches and the estimation of flood flow in some river reaches. Among them, there are records of many river reaches in 1849. “The Great Flood of 1954 in Yangtze River” [56] records and studies the 1954 flood in the Yangtze River Basin, one of the largest floods since the beginning of China’s hydrological instrumental records. The survey data and investigation in this book are momentous for the analysis and evaluation of the flood in 1849.

2.2. Methods

The historical geography textual research method implemented in this study is a common research method in the Chinese historical geography field [57]. Briefly, it systematically collects weather and disaster information recorded in historical documents, and converts the Chinese lunar calendar date at that time into the Gregorian calendar, and also relocates the location names at that time to reconstruct the temporal and spatial processes of weather events.

According to historical documents, this study identified the areas of extreme precipitation. Based on the collection of historical records, the records of heavy rains and floods with clear time records were identified in county units. The counties with records of heavy rains and floods in the same time period and same watershed constitute an extreme precipitation area modified using the distribution of modern rain belts.

Due to limited flood flow rate measurements in historical periods, the reconstruction of flood flow rate mainly relied on vague historical records and flood traces. As a result, the reconstruction of flood flow rate has been one of the most difficult problems in historical hydrology. Many previous studies have discussed it, e.g., simulation using flood level and topographic data [58] and environmental reconstruction to reconstruct the flow and process of flood event of the artificial breach of the Yellow River in Kaifeng in 1642 [59]. In this study, the extreme peak discharges of several river sections of the Yangtze River tributaries were calculated using the surveyed flood water level and river bed cross section extracted from “The Compilation of Survey Data of Historical Great Floods in China” [55]. The peak discharge of the stations in the mainstream of Yangtze River was calculated using the water level–discharge relationship curve in “The Great Flood of 1954 in Yangtze River” [56]. Based on historical documents and existing studies, we performed an assessment on the peak discharge of those stations where riverbed shapes were significantly changed. We reconstructed the process and intensity of the flood peak using historical documents (which contain dam failures and urban flood information along the Yangtze River) and current river bed and water level data. This study also analyzed the characteristics of the flood water level, maximum peak discharge and flood process using current hydrological data and “The Compilation of Survey Data of Historical Great Floods in China” [55] (recording the historical water level and peak discharge of several river reaches in the Yangtze River).

Meiyu is a unique weather phenomenon of the East Asian monsoon region in the early summer. It is generated by a stable rain belt formed between 28° and 33° N with the advancement of the subtropical high from June to July. It has a profound effect on agricultural production, residents’ lives and the occurrences of drought and flood in East Asia. Following Zheng Weiwei’s method [60], this study determined the Meiyu period in 1849 using the number of days of daytime precipitation to set the starting and ending days of the Meiyu period. According to modern precipitation observational principles, daytime precipitation is bounded by 8:00 a.m. and 8:00 p.m. After 25 May each year, if there are 5 days of precipitation (among which at least 3 days of precipitation ≥1 mm) or 4 days of precipitation (daily precipitation ≥1 mm) in 10 consecutive days, this can be considered as a precipitation concentration period. The first (last) rainy day of the first (last) precipitation concentration period is regarded as the starting (ending) day of the Meiyu period. During this period, there must be no more than five consecutive rainless days and no less than four consecutive rainy days, and multiple precipitation concentration periods are allowed.

In addition, this study also adopted the method of historical drought and flood events classification. This method derives drought and flood records from some minor stations of China in a specific year, and grades the drought and flood disaster records to obtain the distribution features of drought and flood events in the entire country. This method has been tested, and has been shown to be effective and successful in “The Atlas of Distribution Droughts and Floods in past 500 years in China” [61]. However, the atlas only provides some rough contour lines. To show the spatial distribution of droughts and floods more intuitively, we used Yang Yuda’s method, which contains more stations and thus makes more detailed distribution patterns [62]. We obtained drought and flood distribution maps of China (showing the summer rainfall situation) covering almost the entire country in the year. This study shows the summer rainfall distributions of 1849, 1931, 1954, and 1998, and the data are from “The Atlas of Distribution Droughts and Floods in past 500 years in China” [61] and “The Sequel Atlas of Distribution Droughts and Floods in past 500 years in China” [63].

3. Results

3.1. Reconstruction of the Temporal and Spatial Process of Summer Precipitation in 1849

In June–July 1848, Hubei, Hunan, Jiangxi, Anhui, Jiangsu, and Zhejiang Provinces were repeatedly affected by heavy rainfall. Specifically, Jiangsu and Zhejiang Provinces suffered from many typhoon events in July [41] (pp. 865–876). All these meteorological events caused severe floods at the middle and lower reaches of the Yangtze River, among which some did not dry up until the next spring.

In spring of the subsequent year, rainfall days still dominated in Hunan, Jiangxi, Anhui and Hubei Provinces, and floods were serious along the riverbank and in low-lying areas. According to a report by Zhao Bingyan (Governor of Hunan Province), the rivers and lakes of Hunan Province overflowed in about eight prefectures and counties in April-May, because of the abundant rain, resulting in submerged embankments and fields [41] (pp. 883–885). There were more than 5–10 rainfall events in Hubei Province in April–May [41] (p. 885). There were also numerous rainfall events in Anhui Province in spring, thus the water level rose much earlier than in common years [41] (p. 890). Fei Kaishou (Governor of Jiangxi Province) also reported plenty of rain in Jiangxi Province in late April, and the low-lying fields and levees along the river were flooded occasionally [41] (p. 881).

Since early summer, there were four consecutive heavy precipitation events at the middle and lower reaches of the Yangtze River, and there were several extreme precipitation centers in each event (Figure 2, and see more details in Appendix A,

Table A1. The extreme precipitation sites of Meiyu period in 1849.

Precipitation Events	Extreme Precipitation Site	Longitude (E)	Latitude (N)
The first precipitation event	Mianyang	113.2	30.3
	Xiantao	113.2	30.3
	Qianjiang	112.5	30.26
	An’qing	115.76	29.78
	Huangmei	115.93	30.09
	Wuhu	118.33	31.33
	Tongling	117.82	30.93
	Fanchang	118.21	31.07
	Tongcheng	116.94	31.04
	Wuwei	117.75	31.3
	Hexian	118.37	31.7
	Xingzi	116.03	29.47
	Wannian	117.08	28.7
	Dehua	115.97	29.71
	Jianchang	116.62	27.56
	De’an	115.75	29.33
	Ruichang	115.65	29.68
	Qingjiang	115.54	28.07
	Taihe	114.88	26.81
	Zhenjiang	119.45	32.2
	Suzhou	120.62	31.32
	Changshu	120.74	31.64
	Wuchang	114.29	30.569
	Nanchang	115.898	28.682
	Jiangning	118.77	32.05
	Suzhou	120.62	31.32
	Songjiang	121.47	31.25
The second precipitation event	Hanyang	114.02	30.57
	Jingmen	112.2	31.04
	Xiaogan	113.91	31
	Hanchuan	113.59	30.63
	Huanggang	114.87	30.44
	Huangmei	115.93	30.09
	Guangji	115.56	29.85
	Jiayu	113.91	29.97
	Jiangling	112.18	30.35
	Huangpi	114.36	30.88
	Qianjiang	112.5	30.26
	Pengze	116.56	29.9
	Ruichang	115.65	29.68
	Hukou	116.23	29.75
	Hexian	118.37	31.7
	Xingzi	116.03	29.47
	Wannian	117.08	28.7
	Dehua	115.97	29.71
	Jianchang	116.62	27.56
	De’an	115.75	29.33
	Ruichang	115.65	29.68
	Qingjiang	115.54	28.07
	Taihe	114.88	26.81
	An’qing	115.76	29.78
	Wuhu	118.33	31.33
	Wuwei	117.75	31.3
	Tongling	117.82	30.93
	Taiping	118.13	30.28
	Fanchang	118.21	31.07
	Tongcheng	116.94	31.04
	Changshu	120.74	31.64
	Zhenjiang	119.45	32.2
	Shaoxing	120.5	30
	Huzhou	120.1	30.86
	Jiaxing	120.76	30.77
	Yanzhou	119.27	29.49
	Wuchang	114.29	30.569
	Nanchang	115.898	28.682
	Hangzhou	120.16	30.27
	Jiangning	118.77	32.05
The third precipitation event	Hanyang	114.02	30.57
	Jingmen	112.2	31.04
	Jingzhou	112.1	30.35
	Qianjiang	112.5	30.26
	Huangmei	115.93	30.09
	An’lu	113.69	31.25
	Huanggang	114.87	30.44
	Jiujiang	115.97	29.71
	Pengze	116.56	29.9
	Hukou	116.23	29.75
	Ruichang	115.65	29.68
	Shangrao	117.97	28.47
	De’an	115.75	29.33
	Ganzhou	114.92	25.85
	Xiangyin	112.87	28.68
	Yuanjiang	112.36	28.83
	Yiyang	112.33	28.6
	Taoyuan	111.47	28.9
	Linxiang	113.42	29.48
	Baling	113.09	29.37
	An’xiang	112.16	29.41
	Fenghuang	109.43	27.92
	Yuanling	110.39	28.46
	Yongsui	109.46	28.59
	Jishou	109.71	28.3
	Wuhu	118.33	31.33
	Wuwei	117.75	31.3
	Tongling	117.82	30.93
	An’qing	115.76	29.78
	Hanshan	118.11	31.7
	Dangtu	118.49	31.55
	Fanchang	118.21	31.07
	Taiping	118.13	30.28
	Huaining	116.63	30.41
	Susong	116.13	30.15
	Wangjiang	116.69	30.12
	Tongcheng	116.94	31.04
	Lujiang	117.29	31.23
	Taihu	116.27	30.42
	Guichi	117.48	30.66
	Changshu	120.74	31.64
	Shaoxing	120.5	30
	Haining	120.69	30.53
	Zhenjiang	119.45	32.2
	Suzhou	120.62	31.32
	Wuchang	114.29	30.569
	Changsha	112.98	28.2
	Nanchang	115.898	28.682
	Hangzhou	120.16	30.27
The fourth precipitation event	Xianning	114.28	29.87
	Huangmei	115.93	30.09
	Puqi	113.85	29.71
	Zhongxiang	112.58	31.17
	Yingcheng	113.6	30.94
	Yunmeng	113.73	31.02
	Jianli	112.9	29.83
	Shishou	112.41	29.73
	Tongshan	114.52	29.6
	Yangxin	115.22	29.83
	Jingzhou	112.1	30.35
	An’xiang	112.16	29.41
	Fenghuang	109.43	27.92
	Yuanling	110.39	28.46
	Xiangyin	112.87	28.68
	Yuanjiang	112.36	28.83
	Yiyang	112.33	28.6
	Baling	113.09	29.37
	Yongsui	109.46	28.59
	Jishou	109.71	28.3
	Taoyuan	111.47	28.9
	Linxiang	113.42	29.48
	Jiujiang	115.97	29.71
	Pengze	116.56	29.9
	Hukou	116.23	29.75
	Ruichang	115.65	29.68
	Shangrao	117.97	28.47
	De’an	115.75	29.33
	Ganzhou	114.92	25.85
	Lujiang	117.29	31.23
	Susong	116.13	30.15
	Taiping	118.13	30.28
	Wuhu	118.33	31.33
	Wuwei	117.75	31.3
	Tongling	117.82	30.93
	An’qing	115.76	29.78
	Hanshan	118.11	31.7
	Dangtu	118.49	31.55
	Fanchang	118.21	31.07
	Huaining	116.63	30.41
	Wangjiang	116.69	30.12
	Tongcheng	116.94	31.04
	Taihu	116.27	30.42
	Guichi	117.48	30.66
	Jianping	119.17	31.14
	Qianshan	116.53	30.62
	Shexian	118.44	29.88
	Xiuning	118.19	29.81
	Qimen	117.7	29.86
	Jixi	118.57	30.07
	Nanling	118.32	30.91
	Xuancheng	118.73	31.95
	Qingyang	117.84	30.64
	Taixian	117.48	30.19
	Chaoxian	117.87	31.62
	Yixian	117.92	29.93
	Jingde	118.53	30.28
	Shucheng	116.94	31.45
	Chuzhou	118.31	32.33
	Laian	118.44	32.44
	Quanjiao	118.27	32.1
	Guangde	119.41	30.89
	Zhenjiang	119.45	32.2
	Suzhou	120.62	31.32
Changshu	120.74	31.64
Huai’an	119.022	33.63
Taicang	121.1	31.45
Yangzhou	119.42	32.39
Shaoxing	120.5	30
Huzhou	120.1	30.86
Jiaxing	120.76	30.77
Yanzhou	119.27	29.49
Haining	120.69	30.53
Fuyang	119.95	30.07
Yuhang	120.3	30.43
Jiashan	120.92	30.84
Haiyan	120.92	30.53
Tongxian	120.54	30.64
Deqing	120.08	30.54
Changxing	119.91	30.01
Yuyao	121.16	30.04
Chunan	119.05	29.61
Tonglu	119.64	29.8
Tongren	109.21	27.73
Songtao	109.18	28.17
Wuchang	114.29	30.569
Nanchang	115.898	28.682
Hangzhou	120.16	30.27
Changsha	112.98	28.2

), which have plenty of historical records of heavy rainfall and severe flooding.

The first precipitation event occurred from 18 May to 30 May 1849, and extreme precipitation concentrated from 25 May to 28 May 1849. There was a center of extreme precipitation in the southeastern province of Hubei (30°–31° N, 112°–114° E). According to Zhao Bingyan’s memorial, from 25 May to 28 May 1849, heavy precipitation occurred in many places in Hubei, the water level in both Yangtze River and Han River rose simultaneously [41] (p. 885). At the same time, there was another extreme precipitation center in the middle of Jiangxi (26°30′–28°30′ N, 115°–117° E), where 14 prefectures and counties had much rain from 20 to 30 May 1849, according to the memorial of Fei Kaishou [41] (p. 885). There was a third center of extreme precipitation at the border of Hubei, Jiangxi, and Anhui (29°–30° N, 115°30′–116°30′ E). Apart from these, there was another east–west distributed center of extreme rainfall in the middle-eastern province of Anhui and southern province of Jiangsu (31°–32° N, 117°–121° E). There was extreme rainfall, “pouring down day and night”, in 11 prefectures and counties in Anhui Province, from 23 May to 27 May 1849, according to “The Transcript of Memorials from Ministry of Defense” [42] (p. 03-3502-073). A similar situation occurred in the Yangtze River Delta from 24 May to 27 May [41] (p. 891) [42] (p. 03-3576-006) [53,64] (p. 3961).

The second precipitation event occurred from 6 June to 18 June 1849, lasting for half a month. There was an extreme precipitation center in Southeastern Hubei (30°–31° N, 112°–115° E). According to the memorial of Tang Shuyi (Governor of Hubei), 14 prefectures and counties of Hubei Province experienced heavy rains from 11 June to 19 June 1849, where floodwaters rushed down from the mountains, and the lakes and rivers were rising [41] (p. 885). Meanwhile, there was a strip-type center of extreme precipitation in the central part of Jiangxi (26°30′–28°30′ N, 115°–117° E). Fei Kaishou reported heavy rains in many places in Jiangxi, from 11 June to 20 June 1849; the low-lying fields along the river were flooded and the dikes were washed down [41] (p. 885). From 11 June to 19 June 1849, there was another belt-shaped heavy precipitation center in the southern Anhui Province (30°–32° N, 117°–119° E), and the stagnant water could not be released in the low-lying fields of the counties in Anhui [41] (p. 890) [42] (p. 04-01-01-0836-013). There was also an extreme precipitation center at the junction of Hubei, Jiangxi, and Anhui (29°–30° N, 115°30′–116°30′ E). In addition, there was an extreme precipitation center in the southern province of Jiangsu and northern province of Zhejiang (29°–32° N, 119°–121° E), which was strip-like distributed from southwest to northeast. Fu Shengxun (Governor of Jiangsu) reported the heavy rains in Jiangsu Province, for days and nights from 12 June to 16 June 1849 [41] (p. 891), and continuous rainfall, with only half-day or moments of exceptions, in Zhejiang Province, from 10 June to 16 June 1849 [41] (p. 893) [51,52].

The third precipitation event occurred from 19 June to 30 June 1849, and the extreme precipitation center gradually moved westward. There was an extreme precipitation center in a southwest to northeast direction in the northern province of Hunan (28°–29°30′ N, 109°30′–113°30′ E). Zhao Bingyan reported that Hunan was flooded due to the rising lake water level and flash floods, and many embankments, fields, and houses were flooded in late June 1849 [41] (p. 893). There was also an extreme precipitation center in the southeastern province of Hubei, which was distributed in an east–west direction (30°–31° N, 112°–115° E). It poured for days in Hubei in late June 1849; hence, the Yangtze River, Han River, and many lakes rose and overflowed simultaneously [41] (p. 882). At the same time, there was a band-shaped extreme precipitation center in the northwestern province of Jiangxi and southern province of Anhui (28°40′–32° N, 115°–119° E). About 7 counties in Jiangxi and 14 counties in Anhui experienced heavy rain after 20 June 1849, and the water level of the Yangtze River was one meter above that in 1848 [41] (pp. 886–887) [42] (p. 04-01-01-0832-052). In addition, the Yangtze River Delta exhibited a strip-shaped extreme precipitation center from southwest to northeast (30°–32° N, 120°–121°30′ E), and many diaries recorded long-lasting rainfall in the Yangtze River Delta from 19 June to 30 June 1849 [51,52].

The fourth precipitation event took place from 1 July to 17 July 1849. The center of extreme precipitation was distributed more westward and northward, with the west border to the northeastern province of Guizhou, and the north border to Huai’an in Jiangsu (33°30′ N, 119° E). In the northeastern province of Guizhou and northwestern province of Hunan (27°–30° N, 109°–115° E), there was a large center of heavy rainfall distributed from southwest to northeast. In early and mid-July 1849, 11 counties in Guizhou and Hunan suffered from heavy rainfall [41] (p. 882) [42] (p. 04-01-01-0832-056). There was a center of extreme precipitation in central Hubei (30°–31° N, 111°–114° E). According to the memorial of Yutai (Governor of Huguang), the flood in Hubei had got no place to accommodate, overflowed the river bank, the rivers and lakes were connected just like seascapes” in early and mid-July 1849 [41] (p. 882). At the same time, there was another belt-shaped center of extreme rainfall in the northwestern province of Jiangxi, southern province of Anhui, and western province of Jiangsu, in a northwest–southwest direction (28°30′–32°30′ N, 115°–120° E). According to Wang Zhi’s report, more than 30 prefectures and counties had heavy rain and mountain torrents in mid-July 1849 in Anhui [41] (pp. 887, 891–892) [42] (p. 04-01-01-0832-052). Moreover, there was also a sub-center of extreme precipitation at the junction of Jiangsu, Zhejiang, and Anhui (28°30′–32° N, 118°–122° E). According to the memorial of Wu Wenrong, heavy rain hit 26 counties in Zhejiang, for days and nights, from 2 July to 14 July 1849, which increased the water levels of the rivers and broke the ditches [41] (p. 893). Heavy rain also occurred in Jiangsu from 5 July to 15 July 1849 [41] (p. 893) [65] (p. 11).

The heavy summer rainfall process in the Yangtze River Basin ended roughly around 20 July 1849 [41] (pp. 881–895) [42] (p. 04-01-01-0832-052) [65] (p. 11). After that, the middle and lower reaches of the Yangtze River entered into dry dog days (the hottest and driest period in a year, which occurs from mid-July to mid-August).

3.2. Reconstruction of the Yangtze River Flood Process in 1849

It was rainy in spring in Hunan, Hubei, Jiangxi, and Anhui, at the middle reaches of the Yangtze River, causing floods in low-lying areas. For example, the heavy precipitation in the northern province of Jiangxi caused rising river water levels and flooding in low-lying areas in late April 1849 [41] (p. 881). The dams were overwhelmed and the fields were flooded in the central province of Hunan [41] (p. 881). The heavy rainfall at the middle reaches was closely related to the flood at the middle and lower reaches of the Yangtze River [66], and would increase the water level of the mainstream simultaneously in the Yangtze River.

According to the memorial of Wang Zhi (Governor of Anhui), on 8 May 1849, “the Huaihe River had not returned to its normal storage capacity from the flood disaster last year (1848), more rain occurred since this spring, causing the water level to rise to the level almost three meters below the peak level last year” [41] (p. 890). Another memorial of Wang Zhi, on 2 August 1849, recorded that the peak water level in the summer of 1849 was one meter above the level in 1848 [41] (p. 896). According to the records of modern hydrological survey data, the peak water level in Anqing station (the provincial capital of Anhui) reached 18.14 m in 1849 (all water levels in our study are relative to the base level at Wusong), which was the highest water level recorded in history [55] (p. 399). Therefore, it can be conjectured that the peak water level of Anqing reached 17 m in 1848, and about 14 m on 8 May 1849. The latter was merely two meters below the current warning water level of 16.08 m at Anqing station [55] (p. 399).

With the occurrence of a stronger precipitation event at the middle reaches of the Yangtze River in late May 1849, the water level of the mainstream and main tributaries of the Yangtze River continued to rise. On 29 May 1849, lakes and rivers in central Hubei were flooded, and many dams were damaged [41] (p. 881). After the start of the Meiyu period on 6 June 1849, the Han River, Dongting Lake water system, and Poyang Lake water system entered flood season one after another, and the mainstream of the Yangtze River was facing tremendous flood pressure. According to the memorial of Tang Shuyi (Governor of Hubei), rivers and lakes in Hubei rose simultaneously, causing flooding in many places from 11 June to 19 July 1849 [41] (p. 885).

In early July 1849, the Yangtze River Basin was ushered in a new round of heavy precipitation, and floods continued to hit the middle reaches [41] (p. 885). Being blocked by the Yangtze River, the Poyang Lake in Jiangxi could not discharge. Hence, large-scale floods hit the entire watershed. In mid-to-late July 1849, with the arrival of the fourth round of extreme rainfall in summer, which was also the last round of heavy rainfall during the Meiyu period, flood peaks in many of the main tributaries of the Yangtze River (e.g., Hanjiang, Dongting Lake water system, and Poyang Lake water system) deluged into the mainstream simultaneously, resulting in the record high water levels at many hydrological stations in the middle reaches of the Yangtze River (as shown in Figure 3). The peak water level of Hankou in Hubei Province, at the middle reaches, reached 28.9 m [55] (p. 289). To prevent the water from flowing back to Wuhan city, the discharge gates had to be closed, causing seascape in the city, with ponding of around 1–4 m [41] (p. 882). Most of the residential areas were flooded in the city of Nanjing at the lower reaches. We speculate that the water level of the Yangtze River in Nanjing probably exceeded 10 m [67].

The water level of the mainstream of the Yangtze River remained high, due to flood water from the upstream and middle reaches after the Meiyu period. According to Tang Shuyi’s memorial on 30 August 1849, the water level of Wuhan city began to drop to about 3 m below the highest water level (i.e., a decline to about 26 m in late August). It was close to today’s warning water level of 26.3 m [41] (p. 887), but still higher than the preparation water level of 25 m [56] (p. 36). The water level in Nanjing dropped in probably around mid-September 1849, which was slower than Wuhan [67].

Figure 3a shows the peak water levels and maximum peak flow rates of some key stations at the middle and lower reaches of the Yangtze River in 1849 (see more details in Appendix A,

Table A2. The records of water levels and flood peak flow rates of the Yangtze River in 1849 from “The Compilation of Survey Data of Historical Great Floods in China” [55].

Yangtze River	Hydrological Station	Location	Water Level (m)	Flood Peak Flow Rate (m3/s)	Degree of Credibility	Organ of Survey	Remarks
Main stream of Yangtze River	Luoshan	Honghu, Hubei	31.40	47,000	high	Changjiang Water Resources Committee	/
	Xindi	Honghu, Hubei	30.98	/	high	Changjiang Water Resources Committee	/
	Jiayu	Jiayu, Hubei	30.47	/	high	Changjiang Water Resources Committee	The water level of Jiayu was 31.47 m in 1954
	Jinkou	Jinkou, Wuchang	30.05	/	high	Changjiang Water Resources Committee	The water level of Jinkou was 30.78 m in 1954
	Wuhanguan of Hankou	Hankou, Wuhan	28.90	65,000	high	Changjiang Water Resources Committee	The water level of Hankou was 26.91 m in 1949, and the catchment area was 1,488,036 km2
	Luoyang	Luoyang, Xinzhou	28.61	/	high	Changjiang Water Resources Committee	The highest water level on records
	Echeng	Echeng	27	/	high	Changjiang Water Resources Committee	The highest water level on records
	Huangshi	Huangshigang, Huangshi	25.8	/	high	Changjiang Water Resources Committee	The highest water level on records
	Wuxue	Wuxue, Guangji	22.66	62,500	high	Changjiang Water Resources Committee	The highest water level on records
	Hukou	Hukou, Jiangxi	21.16	/	high	Changjiang Water Resources Committee	The water level of Hukou was 21.64 m in 1954
	Dongliu	Dongliu	19.45	/	high	Changjiang Water Resources Committee	The highest water level on records, and water level was 18.28 m in 1949, which was the second highest water level on records.
	Anqing	Anqing	18.14	/	high	Changjiang Water Resources Committee	The highest water level on records, and water level was 17.27 m in 1949, which was the second highest water level on records.
	Zongyang	Zongyang town, Zongyang	16.96	/	high	Changjiang Water Resources Committee	The highest water level on records, and water level was 15.84 m in 1949, which was the second highest water level on records.
	Datong	Meigeng town, Guichi	16.14	80,000	high	Changjiang Water Resources Committee	The highest water level on records, and water level was 15.18 m in 1949, which was the second highest water level on records, and the catchment area was 1,705,383 km2.
Branches of Yangtze River	Yinjiang	Eling town, Yinjiang	/	/	high	Water Resources Department of Guizhou	The highest water level on records
	Bayanxiong	Changyan town, Huangbei	/	>1480		Water Resources Department of Hubei	The second highest water level on records, and flood peak was 1480 m3/s, which was the third largest flood peak on records.
	Macheng	Xujia town, Macheng	/	>3530	/	Water Resources Departmentof Hubei	The highest water level on records
	Luotian	Chengguan town, Luotian	/	2650	for reference only	Water Resources Departmentof Hubei	The highest water level in history, and the largest flood peak on records was 4240 m3/s on 14 July 1969
	Yujiapu	Liujia he, Jichun	/	5330	for reference only	Water Resources Departmentof Hubei	The largest flood peak on records.
	Yangaxin	Fushui town, Yangxin	/	6100	for reference only	Water Resources Departmentof Hubei	The highest water level in history, and the largest flood peak on records was 3910 m3/s on 17 June 1959.
	Fenghuang	Tuojiang town, Fenghuang	/	1800–1960	/	Water Resources Departmentof Hunan	The second largest flood peak in history, and the largest flood peak was 11,100 m3/s on 30 June 1974.
	Longtou	Longtou town, Baojing	/	>13,100	/	Water Resources Departmentof Hunan	The largest flood peak in history, and the second largest flood peak on records was 13,100 m3/s in 1927.
	Baojing	Qianling town, Baojing	/	13,700	high	Water Resources Departmentof Hunan	The fifth largest flood peak in history, and the largest flood peak was 11,100 m3/s on 29 June 1954.
	Chenjiahe	Chenjiahe, Zhangjiajie	/	>9000	/	Water Resources Departmentof Hunan	The largest flood peak in history, and the third largest flood peak on records was 9000 m3/s on 5 July 1935.
	Chaoxian	Chaoxian	13.57	/	for reference only	Water Resources Departmentof Anhui	The highest water level on records, and the water level was 13.09 m in 1954.
	Taoxi	Taoxi town, Shucheng	/	573	for reference only	Water Resources Departmentof Anhu	The forth largest flood peak on records, and the largest flood peak was 1250 m3/s on 12 July 1991.
	Kaichengqiao	Kaichengq-iao town, Wuwei	/	/	/	Water Resources Departmentof Anhu	The largest flood peak in history.

Note: the water level was relative to the base level at Wusong.

). The maximum flow rate was mainly based on the peak water level/discharge curve of 1954 [56] (pp. 117–122). Among them, further explanation is needed on the peak flow rate of only 47,000 m³/s at Luoshan station (according to the peak water level/discharge curve in 1954). This might be related to the changes in the cross section of the riverbed, caused by the drastic changes in the middle reaches of the Yangtze River. The water level at Luoshan station was 31.4 m in 1849, which was the fourth highest water level in the history of Luoshan station. In 1954, the warning water level at Luoshan station was 31.5 m. Although the flood level in 1998 was 1.78 m higher than that in 1954, its peak water volume was 11,000 m³/s less than that in 1954. The change in the flood peak water volume at Luoshan station was related to the changes in the river near the new dam at the lower reaches of Luoshan. The mid-channel sand lands of this section of the river developed quickly since 1860, and developed into the mid-channel islands in 1934, which narrowed the cross section of the river [68] (p. 68). The flow cross section near Luoshan station was half as narrow as it was in 1849, causing backwater effect during the flood period and a high water level condition. By combining this factor with historical information, we speculate that the maximum flood peak was about 55,000 m³/s at Luoshan station in 1849.

Figure 3b shows the flood peak flow rates and water levels of the hydrological stations where the Yangtze River system reached the highest water level or flood flow rate in history, in 1849 (see more details in Appendix A, Table A2). The highest water level records in history appeared in hydrometric stations, from Yangluo station to Anqing at the middle reaches, and Yuxihe station downstream. The measurements of Wuhan and Wuxue stations in the middle reaches, and Datong and Nanjing stations downstream also nearly hit their highest water levels or peak flow rates in history. These demonstrate that the Yangtze River flood disaster broke out early, lasted for a long time, and the flood volume was extremely large, which was a rare extreme flood in history. At the same time, the river section with the highest water level or peak flow rate was mainly at the middle and lower reaches of the Yangtze River. The mainstream from Jingzhou to Chenglingji (also called the Jingjiang River) had the greatest flood prevention pressure in normal years (because of the curved channel), as it did not experience much flooding this year. This was related to the concentration of precipitation at the middle and lower reaches in summer in 1849. The flood from Luoshan station to Wuhan was obviously related to the water coming from the Dongting Lake system. The extreme high water levels in Wuxue and Yangluo could be related to the torrential rain in the northern tributaries, such as the Han River in July, and the backwater effect of Poyang Lake. The highest water level in history, in Anqing and Datong, probably appeared after 20 July 1849, which might be related to the flood coming from the Poyang Lake system.

The maximum peak flow rate in history, in the tributaries of the Yangtze River, was related to the extreme rainfall during the Meiyu period in July 1849. For example, the maximum peak flow rate of the Yuanjiang River might be related to the heavy rainfall in early July 1849 in the western province of Hunan and eastern province of Guizhou. Whilst the drainage area of the Youshui River in Baolong covered just less than 10,000 km², the peak flow rate reached a remarkable amount of 13,100 m³/s.

4. Discussion

The amount of precipitation in the Yangtze River Basin in early summer is an important cause of the floods in the Yangtze River. The cause of the extreme flood in the Yangtze River in 1849 was unanimously believed to be the extremely long Meiyu period in early summer. The average length of the Meiyu period in East Asia is 26 days [69]. In our earlier study in 2008, we revealed that the Meiyu period in 1849 lasted 62 days, from 18 May to 18 July, which was the longest record in history [38]. Nevertheless, a study by Yan Chaoqiang suggested that the Meiyu period in 1849 lasted for 39 days (from 11 Jun to 19 July) [39]. In a subsequent study on the transition of the rain belt in Eastern China in 1849, however, Yan proposed that the heavy rainfall of 1849 started on 18 May [40]. The main reason for the divergences was that they were mainly based on the rain reports in the archives, whose resolution was too low to make clear judgments.

In this study, we utilized three weather diaries, with a higher resolution for Zhenjiang, Hangzhou, and Shaoxing at the lower reaches of the Yangtze River, apart from the weather diary in Songjiang, used by Yan Chaoqiang. By implementing the method of Zheng Weiwei on the data, we found that the Meiyu period of 1849 was from 6 June to 17 July, which lasted for 42 days and was 62% longer than the average length (Figure 4, and see more details in Appendix A,

Table A3. The precipitations recorded by diaries in 1849.

The Lunar Calendar		The Gregorian Calendar		Guantingfen’s Diary (Hanghzou)		Sanyuantang’s Diary (Zhenjiang)		Pinglu’s Diary (Shaoxing)		Flood Diary in Ji Yu (Shanghai)
Month	Day	Month	Day	Weather Condition	Grade	Weather Condition	Grade	Weather Condition	Grade	Weather Condition	Grade
4	24	5	16	sunny	0	sunny	0	sunny	0	/	/
4	25	5	17	sunny	0	sunny	0	sunny	0	/	/
4	26	5	18	light rain	1	sunny	0	light rain	1	/	/
4	27	5	19	light rain	1	light rain	1	light rain	1	/	/
4	28	5	20	moderate rain	2	light rain	1	sunny	0	/	/
4	29	5	21	light rain	1	sunny	0	light rain	1	/	/
4	30	5	22	light rain	1	sunny	0	light rain	1	/	/
5	1	5	23	light rain	1	moderate rain	2	light rain	1	/	/
5	2	5	24	light rain	1	light rain	1	light rain	1	/	/
5	3	5	25	rainstrom	3	moderate rain	2	rainstrom	3	/	/
5	4	5	26	light rain	1	moderate rain	2	moderate rain	2	/	/
5	5	5	27	light rain	1	light rain	1	light rain	1	/	/
5	6	5	28	sunny	0	sunny	0	rainstrom	3	/	/
5	7	5	29	sunny	0	sunny	0	sunny	0	/	/
5	8	5	30	sunny	0	sunny	0	sunny	0	/	/
5	9	5	31	sunny	0	sunny	0	sunny	0	/	/
5	10	6	1	sunny	0	sunny	0	sunny	0	/	/
5	11	6	2	sunny	0	sunny	0	sunny	0	/	/
5	12	6	3	light rain	1	sunny	0	light rain	1	/	/
5	13	6	4	sunny	0	sunny	0	sunny	0	/	/
5	14	6	5	sunny	0	sunny	0	sunny	0	/	/
5	15	6	6	light rain	1	moderate rain	2	light rain	1	/	/
5	16	6	7	moderate rain	2	light rain	1	light rain	1	/	/
5	17	6	8	moderate rain	2	sunny	0	sunny	0	/	/
5	18	6	9	moderate rain	2	light rain	1	sunny	0	/	/
5	19	6	10	light rain	1	light rain	1	light rain	1	/	/
5	20	6	11	light rain	1	sunny	0	light rain	1	/	/
5	21	6	12	light rain	1	light rain	1	light rain	1	/	/
5	22	6	13	light rain	1	light rain	1	rainstrom	3	/	/
5	23	6	14	moderate rain	2	sunny	0	light rain	1	/	/
5	24	6	15	light rain	1	light rain	1	light rain	1	/	/
5	25	6	16	light rain	1	rainstrom	3	light rain	1	/	/
5	26	6	17	sunny	0	sunny	0	sunny	0	/	/
5	27	6	18	light rain	1	sunny	0	light rain	1	/	/
5	28	6	19	rainstrom	3	light rain	1	light rain	1	/	/
5	29	6	20	moderate rain	2	light rain	1	moderate rain	2	light rain	1
5	30	6	21	moderate rain	2	sunny	0	light rain	1	light rain	1
6	1	6	22	light rain	1	sunny	0	sunny	0	moderate rain	2
6	2	6	23	light rain	1	moderate rain	2	sunny	0	light rain	1
6	3	6	24	light rain	1	moderate rain	2	light rain	1	light rain	1
6	4	6	25	moderate rain	2	light rain	1	rainstrom	3	sunny	0
6	5	6	26	light rain	1	light rain	1	light rain	1	moderate rain	2
6	6	6	27	light rain	1	moderate rain	2	sunny	0	moderate rain	2
6	7	6	28	rainstrom	3	sunny	0	rainstrom	3	moderate rain	2
6	8	6	29	rainstrom	3	sunny	0	rainstrom	3	light rain	1
6	9	6	30	rainstrom	3	light rain	1	rainstrom	3	rainstrom	3
6	10	7	1	sunny	0	light rain	1	light rain	1	moderate rain	2
6	11	7	2	moderate rain	2	light rain	1	light rain	1	light rain	1
6	12	7	3	rainstrom	3	sunny	0	rainstrom	3	light rain	1
6	13	7	4	moderate rain	2	sunny	0	moderate rain	2	rainstrom	3
6	14	7	5	light rain	1	light rain	1	rainstrom	3	sunny	0
6	15	7	6	rainstrom	3	light rain	1	moderate rain	2	light rain	1
6	16	7	7	moderate rain	2	light rain	1	light rain	1	rainstrom	3
6	17	7	8	moderate rain	2	moderate rain	2	light rain	1	light rain	1
6	18	7	9	light rain	1	rainstrom	3	sunny	0	moderate rain	2
6	19	7	10	rainstrom	3	moderate rain	2	moderate rain	2	light rain	1
6	20	7	11	light rain	1	sunny	0	light rain	1	moderate rain	2
6	21	7	12	light rain	1	light rain	1	moderate rain	2	light rain	1
6	22	7	13	light rain	1	sunny	0	light rain	1	light rain	1
6	23	7	14	moderate rain	2	light rain	1	light rain	1	light rain	1
6	24	7	15	sunny	0	light rain	1	sunny	0	moderate rain	2
6	25	7	16	light rain	1	moderate rain	2	light rain	1	light rain	1
6	26	7	17	light rain	1	sunny	0	light rain	1	light rain	1
6	27	7	18	light rain	1	sunny	0	light rain	1	sunny	0
6	28	7	19	sunny	0	sunny	0	sunny	0	sunny	0
6	29	7	20	sunny	0	sunny	0	sunny	0	sunny	0
6	30	7	21	light rain	1	sunny	0	sunny	0	sunny	0
7	1	7	22	sunny	0	sunny	0	sunny	0	sunny	0
7	2	7	23	sunny	0	sunny	0	sunny	0	sunny	0
7	3	7	24	sunny	0	sunny	0	sunny	0	sunny	0
7	4	7	25	sunny	0	sunny	0	sunny	0	sunny	0
7	5	7	26	sunny	0	sunny	0	sunny	0	sunny	0
7	6	7	27	sunny	0	sunny	0	sunny	0	sunny	0
7	7	7	28	sunny	0	sunny	0	sunny	0	sunny	0
7	8	7	29	sunny	0	light rain	1	sunny	0	sunny	0
7	9	7	30	sunny	0	light rain	1	sunny	0	sunny	0
7	10	7	31	sunny	0	sunny	0	light rain	1	sunny	0

). Since the heavy rainfall period that began on 18 May was broken before 10 June, it was not classified as part of the Meiyu period (according to the definition criterion).

Many preliminary studies have investigated the types of summer precipitation in China [70,71,72]. The earlier Meiyu period at the middle and lower reaches of the Yangtze River reflects the strong subtropical high, while the long-time stagnation of the Meiyu belt in the Yangtze River Basin indicates the weak summer monsoon and unusually southerly subtropical high ridge. At the same time, the stagnation of the Meiyu rain belt in June and July indicates the blocking of high pressure in the mid-high latitudes of the Northern Hemisphere, the significance of the westerly wind branch, and the southerly distribution of the south branch of the westerly wind, compared to normal years [32,33]. In mid-to-late July, the blocking effect of the high pressure was lifted, the rain belt began to leap forward to North China, and heavy rainfall formed from the Haihe to Liaohe River Basins.

The reasons for the excessively long Meiyu period and strong precipitation in 1949 need to be further explored. Generally speaking, the extended period of stagnation of the Meiyu rain belt in the Yangtze River Basin indicates the weak summer monsoon and the unusually southerly location of the subtropical high ridge. The key driver of the long Meiyu period in 1998 was the strong El Niño event [32,33,34]. Nonetheless, there were no strong El Niño incidents in 1954 and 1849. In 1849, the temperature in East Asia was abnormal, the middle and lower reaches of the Yangtze River were rainy, and there were more low-temperature weather events in spring. According to Wu Wenrong’s memorial, “the temperature is low, so the flooded seedlings in the low field have been hopelessly damaged, the seedlings in high fields which have not been flooded are also difficult to grow” [26] (p. 894). At the same time, the cold air in North China was also very active, and frost and even snow events occurred in Shaanxi and Henan at the end of September and early October. All these indicate the relatively active cold air in 1849, the more developed blocking of high pressure in the mid–high latitudes of the Northern Hemisphere, and the blocked northward movement of the summer monsoon. In addition, the southwest monsoon was more active and the zonal airflow was strong in 1849. The rainy season in Yunnan started on 8 May, about 15 days earlier than normal [73]. A more active southwest monsoon brought more moisture and stronger zonal airflow. The study by Wang Zhiyi [74] suggested that the intensity of the zonal transport of moisture in the southwest had a significant correlation with the precipitation amount in the western region of Jianghuai during the Meiyu period.

Although the extreme flood was caused by the extremely long Meiyu period (as shown in

Table 2. Features of Meiyu in several great flood years.

Scheme	1951–2000			1849	1931	1954	1998
	The Earliest/Longest/Maximum (The Year of Appearance)	The Latest/Shortest/Minimum (The Year of Appearance)	Average
Starting day (day/month)	19/5 (1991)	9/7 (1982)	16/6	6/6	13/6	12/6	23/6
Ending day (day/month)	16/6 (1978)	1/8 (1954)	12/7	17/7	30/7	1/8	1/8
Total days (day)	58 (1991)	7 (1978)	26	42	47	50	39
Rainfall (mm)	745 (1954)	61 (1978)	246	/	553	745	497

Note: the data are from the study of Quansheng et al. [69], Jingyun et al. [75], and National Climate Center [76].

), similarly to two other great floods in the Yangtze River (1954, 1998), since the instrumental records became available, they showed different characteristics in terms of the spatial pattern of precipitation and flood peaks. Figure 5 displays the summer rainfall conditions of China in 1849, 1931, 1954, and 1998. The figure shows that the precipitation was generally higher in the whole Yangtze River Basin in 1954 and 1998, yet mainly concentrated in the middle and lower reaches in 1849. The nationwide flood year occurred in 1931, when the entire Jianghuai region and the Yellow River Basin received more precipitation, but the flood peaks of the Yangtze River in 1931 were weaker than 1849, as shown in the flood survey data [55]. Apparently, the precipitation center was distributed more northerly in 1931.

There were three precipitation patterns in the four years of 1849, 1931, 1954, and 1998. The influence of different precipitation patterns on the Yangtze River flood peak can be discussed further. The inflow from the upstream basin above Yichang accounted for 54% of the Yangtze River’s flow from May to October in normal years [26] (pp. 6–7), so the inflow from the upstream reaches had a huge impact on the flood peaks in the middle and lower reaches. According to

Table 3. Comparison of flood peak flow rates of four flood years in the Yangtze River.

Scheme	1849		1931		1954		1998
	Highest Peak Water Level (m)	Maximum Flow Rate (m3/s)	Highest Peak Water level (m)	Maximum Flow Rate (m3/s)	Highest Peak Water Level (m)	Maximum Flow Rate (m3/s)	Highest Peak Water Level (m)	Maximum Flow Rate (m3/s)
Yichang	53	47,000	/	/	55.73	66,800	54.50	66,300
Luoshan	31.40	65,000	31.85	43,000	33.17	78,800	34.95	68,800
Hankou	28.90	65,000	/	/	29.73	76,100	29.43	71,000
Jiujiang	21.60	70,000	20.16	65,000	22.08	73,000	23.03	73,100
Datong	16.14	80,000	15.02	67,000	16.64	92,600	16.32	82,300

Note: The data of 1954 and 1998 are from “The Records of Yangtze River· Flood Control” [45] (p. 466), and the water level data of Yichang Station are estimated data. The survey data of nearby Hukou station were adopted as the water level data of Jiujiang Station in 1849 and 1931. The 1849 flow data are calculated from the 1954 water level/flow curve [56] (pp. 117–122). Since the flow data at Jiujiang Station in 1849 and 1931 are inaccessible, the estimated data are adopted instead.

, the inflow of water above Yichang in 1849 was less than those in 1954 and 1998. Therefore, the floods in the Jingjiang section (where the flood control pressure was the heaviest) were less severe. However, the Yangtze River mainstream from Yangluo station to Poyang and Hukou showed the highest water level in history (Figure 3), which exceeded the great floods of 1954 and 1998. The flood peaks of hydrological stations such as Jiujiang and Datong in 1849 were slightly lower than those in 1954 and 1998, probably due to the flood diversion effect of a large number of dams in the middle reaches of the mainstream and tributaries at that time. There were numerous flooded areas in the middle reaches (i.e., 56 out of all 59 counties in Anhui, 33 out of 68 counties in Hubei, 39 out of 76 counties in Hunan, and 25 out of 79 counties in Jiangxi) [38]. Despite the significantly less inflow from the reach above Yichang in 1849, relative to 1954 and 1998, the runoff caused by extreme rainstorms in the middle reaches was significantly greater. The flood caused by extreme rainfall in the middle reaches of the Yangtze River requires more attention.

Since 1849, the Jingjiang section of the middle reaches of the Yangtze River (where the flood control pressure was the heaviest) rushed south and formed the Ouchi River. It also rushed south and formed the Songzi River in 1860, in the upstream river, for the Yangtze River to flood into the Dongting Lake [77] (pp. 196–298). Taking the year 1954 as an example, the peak flow rate of the Ouchi River diversion reached 14,800 m³/s, and the peak flow rate of Songzi River reached 10,140 m³/s. These accounted for 37.3% of the discharge of incoming water above Yichang [78]. In 2006, the Three Gorges Reservoir on the Yangtze River started to operate, with a flood control capacity of 221.5 × 10⁸ m³. This played an important role in regulating the flood peak from upstream. In 2020, the maximum inflow peak from upstream to the Three Gorges Reservoir reached 75,000 m³/s, and the outflow from the reservoir was only 49,200 m³/s. With a peak cut of 34.4%, it effectively relieved the downstream flood pressure [79]. It also relieved the flood control pressure from Yichang to Luoshan station, to a certain extent.

However, the flood control pressure brought by heavy rains at the middle and lower reaches, similar to that in 1849, should not be underestimated. The reservoir capacity of embankment systems at the middle and lower reaches of the Yangtze River have been greatly improved, compared to those in 1954 and 1998. For example, the safety guarantee water levels at Anqing station and Datong station were raised to 19.34 m and 17.10 m, respectively, i.e., much higher than the flood level in 1954 [45] (p. 459). In recent years, the flood diversion capacity of lakes and wetlands along the mainstream of the Yangtze River (e.g., Dongting Lake and Poyang Lake) has gradually weakened. The area of Dongting Lake dropped by 1725 km² from 1949 to 1995, and the volume decreased by 126 × 10⁸ m³, accordingly [80]. At the same time, with the socioeconomic development at the middle and lower reaches of the Yangtze River, the cost of using wetlands to divert floods is increasing. Therefore, the flood disaster type of the Yangtze River, caused by extreme precipitation at the middle reaches, similar to that in 1849, would result in a significant loss of life and property, and requires more attention.

Due to some ambiguity in historical documents, the spatial extent and time course of precipitation events were not clearly recorded. Also, the survey data in historical periods might not be equivalent to the measured data in modern times. The reconstruction methods of precipitation and flood processes in historical periods would inevitably contain some uncertainties. As for our study, the reconstruction of the spatial extent of extreme precipitation might bring uncertainty, due to the incomplete historical documentation and limited information on the flood process in the Yangtze River (only contained the flood peak height for the main hydrological stations). However, given the various sources of historical documents and survey data in 1849, the uncertainty of our reconstruction work should be restricted to some specific quantitative estimates, which should have little influence on our qualitative judgment about the characteristics of precipitation and flood processes.

5. Conclusions

Based on archives, diaries, and historical flood survey data, this study reconstructed the temporal and spatial patterns of precipitation and hydrological processes of the great flood disaster in the Yangtze River Basin in 1849.

In 1849, precipitation started earlier than in normal years, with four obvious precipitation events since May. The Meiyu period lasted for 43 days, and there was more precipitation.

The precipitation showed a typical belt-like distribution in the Yangtze River Basin in 1849. There were at least two precipitation centers, concentrated separately at the Taihu Lake Basin in the east, and from the eastern province of Hubei to midwestern province of Hunan in the west. Especially, the precipitation center in the west had a huge impact on the Yangtze River flood in the same year.

The cold air was active in the spring and early summer of 1849, and the southeast monsoon was weak. These caused the rain belt to stagnate in the Yangtze River Basin, for a long time. The onset of the southwest monsoon was earlier and more active, which supplied abundant moisture to the Yangtze River Basin. These were the meteorological causes for the flood disaster in 1849.

The Yangtze River flood season in 1849 started earlier and lasted longer than normal years. It entered the flood season in April, and the flood did not retreat until mid-September. Some stations along the mainstream in the middle reaches reached their highest water levels in history. The flood peak was primarily caused by extreme precipitation at the middle reaches. At present, the water conservancy projects on the Yangtze River are mainly built for floods caused by heavy rains in the upper reaches. In the future, the risk of floods in the Yangtze River would increase with the increase in extreme precipitation and the influences of human activities. Therefore, the pattern of precipitation in the middle reaches of the Yangtze River should receive greater attention in future flood control.