黄河下游河道改道的地理变化特征_英文_

黄河下游河道改道的地理变化特征_英文_ J. Geogr. Sci. 2011, 21(6): 1019-1036 DOI: 10.1007/s11442-011-0897-7 ? 2011 Science Press Springer-Verlag The geo-pattern of course shifts of the Lower Yellow River 1*1,2WANG Yingjie, SU Yanjun 1. State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, CAS, Beijing 100101, China; 2. Graduate University of Chinese Academy of Sciences, Beijing 100049, China Abstract: The changing pattern of the Lower Yellow River (LYR) obtained from the traditional studies, which mainly did literal analysis based on historical documents related to the LYR are too macroscopic and absent of intuitiveness. This paper integrates all the records in historical documents related to course shift, flood and overflow of the last 3000 years and stores them in a GIS database. Then, all the data will be visualized in the form of map, which is helpful to show and understand the rules those events conform more intuitively and accurately. Taking these data as foundation, this study summarizes characteristics of the LYR’s courses and influence scope, and classifies them both into three types; divides the flow directions of the LYR’s courses into two periods, and proposes its changing pattern; concludes the character- istic of diversion points of courses shift events; calculates the velocity of courses shifts, gra- dient and sinuosity, and analyzes their changing patterns. Finally, this study classifies factors that may influence the occurrence of a course shift into two types: the internal factors, such as sediment rate, gradient and sinuosity of the river, and the external factors, such as precipita- tion and human activities. Keywords: Lower Yellow River; course shift; visualization; geo-pattern 1 Introduction The LYR has been characterized as frequent course shift, flood, and overflow in history. According to the records in historical documents, there were over 1500 times of breaching and overflowing events, and at least 26 times of large course shift events during the last 3000 years (Shen et al., 1935). These changes have brought unimaginable catastrophes to the people who lived along: thousands of people died, and numerous people were in the state of homeless. Thus, it is highly beneficial to learn the changing pattern of the LYR, which can be used to forecast the occurrence of the next disaster and minimalize the loss. Further- more, understanding these principles can help the government implement more effective Received: 2011-05-10 Accepted: 2011-06-16 Foundation: Project supported by the Ministry of Science and Technology of China, No.2009BAH50B01; No.2008BAK50B05 Author: Wang Yingjie (1961–), Professor, specialized in spatial information visualization. E-mail: wangyj@igsnrr.ac.cn *Corresponding author: Su Yanjun, graduate student, E-mail: suyanjun1987@gmail.com www.geogsci.com springerlink.com/content/1009-637X policies to manage the LYR. Many studies have focused on course shift of the LYR. Shen et al. (1935) collected all the records of the Yellow River in historical documents, which were arranged in chronological order and then used to produce maps of paleochannels. YRCC (1984) collected more accu- rate and detailed history of the Yellow River‘s water conservancy of different dynasties, and arranged them into dependent tables in chronological order. The chronology of the Yellow was perfected by improving the precision of time and space information, and the severe plagues of flood and events of course shift were summarized in one table, which provided foundations for the further exploration about the rules of course shift (Yao, 1987; Xu H L, 1993). Based on these works, ECCPG-CAS (1982) thought that the flow direction of LYR shifted first from north to east, and then from east to south. Wang et al. argued that there were three main forms of directions of the LYR—flowing northward, eastward and south- ward, and gave a more detailed and accurate description about the changing pattern of course shift. (Wang and Yang, 1993b). Although these results provided a macroscopic depic- tion about the course shift rules, due to the limitation of literary records, these studies could not provide a precise result of the changing pattern of the LYR, which were absent of details. Additionally, the results obtained in these studies were in literal form, which were also ab- sent details and intuitiveness. The Geography Information System (GIS) and Remote Sensing (RS) are useful tools to compensate the disadvantage of non-visualization of literary records, and are widely applied in the study of other rivers. Downward et al. (1994) made a fairly complete description about method of studying river changes with the help of antique maps and aerial photo- graphs. Changes of Dee River, including course shift, lateral shift of the channel midpoint, changes in channel width, changes in the pattern of sediment deposition, changes in bank vegetation, changes in sinuosity and so on, were directly visualized and analyzed based on antique maps and aerial photographs (Gurnell, 1994; 1997; Leys and Werritty, 1999). Goswami et al. (1999), Sarama et al. (2007), and Das et al. (2007) studied the changes of Subansiri, Brahmaputra in Assam India and Barak River, in N.E. India, respectively. Buck- ingham et al. (2007) focused on quantifying channel changes in Las Vegas, Nevada with RS images in different periods. Recently, this method is also used to the study of YR. Zhao et al. (1996) monitored the occurrence of ice-jam flood in the LYR using RS images. Yang et al. (1999) selected the MSS and TM data from 1976 to 1994 covered the LYR to study the changes in the LYR channels. Chang et al. (2004), Huang et al. (2004), Cui et al. (2007), Xue et al. (2009), and Cui et al. (2010) utilized RS images in different periods to study the changing pattern around the estuary of the Yellow River. Maps, aerial photographs and RS images provide direct visualization and geographic at- tributes of the LYR at a particular time, and comparison and analysis among them are ex- pected to produce more accurate and reliable results. However, the fact that GIS maps and RS images are only available in specific but short period, which prevents the study on the overall pattern of course shift of the LYR with a long history about thousands of years. Thus, it is a critical but difficult problem about how to convert historical documents into data that can replace maps, aerial photographs and RS images to study the geo-pattern of course shift of the LYR. This paper presents a way to transform historical documents to visual maps using GIS to study the pattern of courses shifts of the LYR. The historical documents were first trans- WANG Yingjie et al.: The geo-pattern of course shifts of the Lower Yellow River 1021 ferred into vector that could be used in GIS software; then, the course shifts were visualized, and their geo-pattern was extracted; finally, the factors that affected the course shift were discussed. 2 Data and methods 2.1 Study area In history, to the north, the LYR has ever captured the Haihe River and run into the Bohai Sea at Tianjin. To the south, it has ever captured the Huaihe River and run into the Yellow Sea. The entire basin of the LYR, mainly located on the North China Plain (32?–40?N, 2114?–121?E), covers an area of 310,000 km, including Beijing, Tianjin, Hebei Province, Shandong Province, Henan Province, Anhui Province and Jiangsu Province (Figure 1). The terrain of this area is flat, and the average altitude is about 50 m, which belongs to a typical alluvial plain. Nowadays, the LYR empties into the Bohai Sea at Lijin, Shandong Province, and the starting point of the present LYR is Taohuayu in Henan Province (Figure 1). For the purpose of fully using historical documents, this paper takes Yongji County as the unified starting point to study the changing pattern of the LYR. The LYR is famous for its heavy sediments load, because the middle Yellow River runs through the Loess Plateau with fragile ecologi- cal environment and serious soil erosion. The average annual runoff of Huayuankou (in 3Zhengzhou, Henan Province) is 47 billion m, while the average sediment discharge reaches Figure 1 Map of the study area up to 16 billion (Wang and Yang, 1993a). This heavy silting continues to raise the riverbed of the LYR and the LYR is then gradually growing into a suspending river. Nowadays, the suspending situation of the LYR seems to be more serious than ever before. The highest point of riverbed exceeds peripheral ground 10 m in Kaifeng City, and an average height of 4–6 m from Zhengzhou to its outlet (Huang et al., 1992). Once the middle Yellow River is in flood, the LYR has a great possibility to breach and then shift its course. 2.2 Data preprocessing This study used two kinds of data including historical documents and digital data to estab- lish the spatio-temporal database of flood, overflow, and course shift events of the LYR. The source and preprocessing flow of each kind of data is discussed as follows. 1) Historical documents The flood, overflow, and course shift events of the LYR have been recorded in historical documents in different periods, such as Shiji, Hanshu, Houhanshu, Jinshu, Jiutangshu, and so on. For example, Shiji written by a famous Chinese historian and litterateur named Si Maqian elaborated important events and changes of the Yellow River occurred from 2697 B.C. to 122 B.C. Hanshu, Houhanshu, Jinshu, Jiutangshu, and so on recorded significant events happened in corresponding dynasties in Chinese history. Later scholars have under- taken a lot of work on the arrangement of these materials. The Chronology of Yellow River (Shen et al., 1935) and Summary History of Yellow River Water Conservancy (YRCC, 1984) are two outstanding examples of such work. The Chronology of Yellow River collected all records related to Yellow River, classified them into flood, overflow, course shift, repairing, policy modifying events and so on, and arranged them into 7 tables. Summary History of Yellow River Water Conservancy collated the breaching and overflowing events of the LYR and arranged them in the chronological order of different dynasties. In our study, 1573 breaching and overflowing events of the LYR and 50 course shift events collected from the two documents on the Yellow River were used to establish a spa- tio-temporal database. It should be noted that the spatial and temporal information of some events were missed, which cannot be used to establish the database and analyze the chang- ing pattern of the LYR. Finally, this study selected out 1043 breaching and overflowing events and retained all the 50 course shift events. Additionally, historical documents men- tioned earlier were referred to validate the reliability of the two documents on the Yellow River and checkup spatial and temporal information of every event. The spatial information in the form of old geo-names and most of them are no longer used nowadays. Thus, the spatial information cannot be used to fix position directly. Thus, a one-to-one correspondence table between old geo-names and present geo-names should be created using Encyclopedic Dictionary of Ancient and Modern Chinese Geographical Names (Dai et al., 2005) that was of great authority and covered almost all the old names which can be commonly seen in Chinese history. Additionally, the chronicles of different provinces were another source helpful to establish this correspondence. 2) Digital data The point and polygon maps of county administrative divisions of China at a scale of 1:1,000,000 are obtained from National Geomatics Center of China (Figure 2), and are mainly used for locating the spatial information which has been turned into nowadays geo-name. Besides, this study also refers to the Shuttle Radar Topography Mission Digital WANG Yingjie et al.: The geo-pattern of course shifts of the Lower Yellow River 1023 Elevation Models (DEM) at a resolution 90 m (Figure 3), which is used to understand the terrain of the study area and get the altitude value of some specific points. In fact, it would be more accurate to use different DEMs corresponding to eras of different courses. However, it is impossible to reconstruct the topography of the study area hundreds even thousands years ago. Thus, this study has to use the current DEM to represent ancient topography. Figure 2 County administrative map of the study area Figure 3 SRTM DEM data of the study area 2.3 Methods 1) Visualization Time information is generally visualized in two forms: event and state (Wang et al., 2003). An object in its lifespan has different states, and events are defined as processes that trans- form it form one state to another. In this study, the changes of the LYR were visualized in the maps using course shift events that meant formation of new courses. However, it would be too overcrowded to visualize all the course shifts in a map. To solve this problem, this study first selected 6 major course shifts. The rules of judging a major course shift are defined as follows: the course shift shares large proportion of the river under study, and should have at least 1 fixed bed; the river remains stable for a long time after the course shift; characteristics of the new course significantly differ from those of the previous courses. Based on the occurring time of these 6 major course shifts, this study divided the last 3000 years into 7 periods, and then visualized all course shifts in each period, respec- tively 2) Profile chart of course shifts As shown in Figure 4, two concentric arcs were created with the same vertex located in Taohuayu to get two profiles. The radius of arcs was determined according to the principle of creating intersecting points as many as possible between arcs and courses to reflect char- acteristics of more courses. Then, angles from these intersecting points to east were calcu- lated (the angle clockwise from the base line was defined as positive). The year of each course shift was set as X axis, and all the angles were plotted on a chart. In this chart, the changes in flow direction were represented by lines between every two points adjacent in time. The line in blue represents the flow direction of a course shift to south, and that in red represents the flow direction of a course shift to north (Figure 5). Figure 4 A schematic diagram for obtaining profiles * The negative number on the X axis represents year before Christ; ―EF‖ represents ―Eastward flow‖. Additionally, the staring time of Yuhe River is represented by1000 B.C. for the reason that the exact origination time is unknown. Figure 5 The profile chart of the main courses 3) Parameters The Annual Changing Speed (ACS), Gradient and sinuosity were the parameters defined to analyze the changing pattern of the LYR. The ACS reflects the average changing speed between two course shifts, which can be calculated from Eq. (1): (1) ω = ? / LT where ω is the ACS; ? is the difference between two vertex angles of two adjacent course shifts; LT is the total life time of corresponding types of flows. The gradient of a course can be calculated in Eq. (2): (2) G = ?H/S where G is the gradient; ?H is the drop height from the highest point to the estuary of the course, which is acquired from DEM data; S is the distance between them, which is calcu- lated from the maps of courses. The sinuosity of a course can be represented by buckling rate, which is calculated in Eq. (3): WANG Yingjie et al.: The geo-pattern of course shifts of the Lower Yellow River 1025 ˆ ˆ W = AA/ AA(3) ˆ ˆ where W is the buckling rate of the river, AAis the length of the curve between the start point and estuary, and AA is the linear distance between them. 3 Results 3.1 Types of courses This study selected the course shifts in 602 B.C., 11 A.D., 1048 A.D., 1194 A.D., 1494 A.D., and 1855 A.D. as the 6 major course shifts, and divided last 3000 years into 7 periods: be- fore 602 B.C., 602 B.C. to 11 A.D., 11 A.D. to 1048 A.D., 1048 A.D. to 1194 A.D., 1194 A.D. to 1494 A.D., 1494 A.D. to 1855 A.D. and after 1855 A.D. Course shifts in each period were respectively visualized in Figure 5. As can be seen, based on the flow direction, these courses were divided into 3 types: the northward flow, the eastward flow and the southward flow. 1) The northward flow. Figures 6a, 6b and 6d showed that northerly courses flowed across the North China Plain, emptied into the Haihe River, and entered the Bohai Bay. They could be easily divided into two periods: one was from the Yu River to 11 A.D. (Figures 6a and 6b); the other was from 1048 A.D. to 1194 A.D. (Figure 6d). Due to the fact that it is difficult to determine the start time of the Yu River, the exact duration of the northerly course is un- known. 2) The eastward flow. As shown in Figures 6c and 6g, the easterly courses flowed across the nowadays Shadong Province and entered the sea between Bohai Bay and Laizhou Bay. This kind of courses also existed in two periods in history: one from 11 A.D. to 1048 A.D. (Figure 6c) and the other from 1855 A.D. to now (Figure 6g). The total duration of easterly course is 1283 years, which is still increasing. 3) The southward flow. As can be seen in Figures 6e and 6f, the LYR captured the Sishui, Suishui, Guohe and Yingshui rivers, entered the Huaihe River, and finally empted into the Huanghai Sea from 1194 A.D. to 1855 A.D. Courses were very unstable and frequently shifted in this period, as shown in Table 1. Generally, courses in this period can be also di- vided into three parts: before 1494 A.D., the LYR‘s main channel flowed southeast to Sishui River; between 1494 A.D. and 1642 A.D., the LYR mainly went east-southward to Guohe and Yingshui rivers; after 1642 A.D., the main channel of LYR returned to flow to southeast to Sishui River, and frequently shifted to capture the Daqing River. 3.2 The influence scope of LYR As seen in Figure 7, the LYR‘s courses swayed within a sector area with the vertex called Taohuayu. The course had ever flowed into the Haihe River to the upmost north, and had once captured the Huaihe River to the upmost south. The influence scope of the LYR in- cluded Shanxi Province, Henan Province, Shandong Province, Hebei Province, Anhui Prov- 2 ince and Jiangsu Province with an area reaching over 310,000 kmthat almost amounted to the total area of France. *In Figures 6e and 6f, the Huaihe River, Sihe River, Guohe River, Yinghe River, Suihe River and Daqing River indicate that the LYR once captured their courses. Figure 6 Maps of course shifts WANG Yingjie et al.: The geo-pattern of course shifts of the Lower Yellow River 1027 Table 1 Course shifts of the LYR from 1194 A.D. to 1855 A.D. Index Start time End time River captured Index Start time End time River captured 1 1194 1494 Sishui River 16 1565 1567 Sishui River 2 1288 1310 Guohe River 17 1567 1596 Sishui River 3 1368 1437 Sishui River 18 1596 1642 Sishui River 4 1391 1437 Yingshui River 19 1601 1603 Huihe River 5 1411 1431 Wenshui River 20 1642 1643 Guohe River 6 1416 1438 Guohe River 21 1644 1855 Sishui River 1453 Guohe River 22 1721 1722 Daqing River 7 1448 1453 Yingshui River 23 1722 1723 Daqing River 1558 Guohe River 24 1751 1751 Daqing River 8 1489 1547 Yingshui River 25 1761 1761 Daqing River 1508 Sishui River 26 1778 1780 Guohe River 1505 Guohe River 27 1781 1785 Daqing River 1505 Yingshui River 28 1787 1787 Guohe River 9 1492 1495 Sishui River 29 1797 1797 Guohe River 1558 Suishui River 30 1803 1804 Daqing River 10 1505 1508 Sishui River 31 1813 1815 Guohe River 11 1509 1530 Sishui River 32 1819 1820 Guohe River 12 1530 1565 Sishui River 33 1841 1841 Guohe River 13 1534 1558 Guohe River 34 1843 1844 Guohe River 14 1540 1542 Guohe River 35 1855 — Daqing River 15 1558 1565 Sishui River — — — — *The ―—‖ represents that the value of it is unknown or ambiguous. Additionally, the basin of LYR was also divided into 3 types: northward flow basin, east- ward flow basin and southward flow basin (Figure 7). The boundary between northward flow basin and eastward flow basin was the radial toward northeast, and that between east- ward flow basin and southward flow basin was the radial toward east. The northward flow basin mainly located within the boundary of Henan and Hebei Province, eastward flow basin within the boundary Henan and Shandong Province, and southward flow basin within the boundary of Henan, Anhui and Jiangsu Province. 3.3 Changes in the velocity of course shifts As shown in Figure 6 and Table 1, the course shift events occurred more and more fre- quently as time went by. However, this may result from the fact that events of the Yellow River were recorded more frequently and in more detail. To study changes of the rate of course shifts, this paper calculated the ACS between every two major course shifts (Table 2). The ACS is not influenced by records, because once the vertex angle and lifetime of a course are decided, it is a constant. It is worth noting that the ACS of the first major course shift and the sixth major course shift is not calculated, because the origin of the Yu River remain un- known and the present LYR has not shifted yet. Figure 7 The division of the influence scope of LYR Table 2 The ACS between every two major course shifts of the LYR –1Index Type Vertex angle Lifetime (Year) ACS (Year) ———flow 1 Northward 0.417 —— 2 Northward flow –30.236 613 0.38×10 3 Eastward flow –3 –0.534 1037 0.52×10 4 Northward flow –3 1.299 146 8.90×10 5 Southward flow –30.669 300 2.23×10 6 Southward flow –3 –0.881 361 2.44×10 7 Eastward flow ———Table 2 indicated that the shifting velocity of the LYR grew rapidly over time, and that of southward flow was much higher than eastward flow and northward flow. The ACS between the fourth and fifth major course shifts was an exception for the reason that the fourth major course shift was largely influenced by human activities. The high sediment rate may be re- sponsible for the more and more unstable LYR. After each course shift, the LYR tended to capture other rivers‘ courses. The high sediment rate resulted large amounts of silt deposited on and then increased the height of riverbed, which made the river more unstable. 3.4 Changes in flow directions As is mentioned, courses of the LYR swayed in a sector. However, the movement of the WANG Yingjie et al.: The geo-pattern of course shifts of the Lower Yellow River 1029 courses is not disorderly. On the contrary, as shown in Figure 5, they took a periodic move- ment in this sector. Changes of flow direction were easily divided into two periods: the period shifting to south (from the Yu River to 1494 A.D.) and that shifting to north (from 1494 A.D. to now). As shown in Figure 8a, the LYR kept shifting to south in the first period until it reached the southernmost boundary of the influence scope of the LYR. After 1494 A.D., courses began to shift to north, and finally formed the present LYR (Figure 8b). Some exceptions to this pattern were also found such as the course shift to north in 1048 A.D. However, most of these course shifts were results of human activities, and river courses formed after these shifts were generally quite unstable and lasted a very short time. For instance, the new courses formed after the major course shift in 1048 A.D. frequently shifted to south, and ?3 ?1only lived for 146 years. As shown in Table 2, the ACS in this period reached 8.90×10(Y) that was nearly 9 times that of the average ACS of other periods. Thus, the overall changing pattern of flow directions of the LYR was summarily de- scribed as follows: it first flowed toward the north by east from the Yu River to 1494 A.D, then turned to flow toward east by north, east by south, south by east in order; after it reached the southernmost boundary of the sector, the river tended to flow toward east by south again, and finally flowed east by north to form the present LYR. If there is a course shift in the future, the new courses have a great possibility to shift to north from the present Yellow River. 3.5 Changes in diversion points Diversion point is generally defined as the place where a river changes its flow direction. Figure 9 shows that diversion points of the LYR do not scatter randomly as they seemed; on the contrary, most of them are clustered in a triangle with the vertexes named Taohuayu, Puyang and Tongwaxiang. All the 6 diversion points of major course shifts and 50% (22 of 44) of the diversion points of secondary course shifts are in this triangle. While the triangle 2 is only about 8,445 kmthat is 2.9% of the influence scope of the LYR. The reason for this pattern may be related to changes of gradient. The average ratio of the gradient of a course before and after the diversion point is calculated as 5.62:1. The big dif- ference in gradient may result in a sudden but sharp decrease of the flow speed after the di- version point, and a large amount of silt deposit around the diversion point, which leads the course unstable and prone to shift. However, this is only one possible explanation for this phenomenon that may be also related to sinuosity, topography, precipitation and so on. 3.6 Changes in gradient and sinuosity The gradient and sinuosity of main courses which the LYR has ever flowed through is first calculated. Table 3 indicates that the gradient of the LYR has changed significantly with the highest gradient of 16.6%, larger than the lowest. The duration of a river decreases with the gradient (courses between 1048 A.D. and 1194 A.D. were exceptions due to the fact that they were forced to flow toward north again due to the influence of human activities). The average gradient of easterly flow is 0.338‰ that is the highest among the three types of flow, and it has lasted 1283 years, which is still increasing; the average gradient of northerly flow is 0.328‰, and courses have lasted at least 759 years; the average gradient of southerly Figure 8 Changes in flow direction (a. the period before 1494 A.D., and b. the period from 1494 A.D. to now) WANG Yingjie et al.: The geo-pattern of course shifts of the Lower Yellow River 1031 Figure 9 Changes in diversion points flow is only 0.301‰ which is the smallest among the three, and its duration is only 661 years that is also the shortest. This study used the buckling rate to represent the extent of sinuosity. As seen in Table 3, the duration of a course seems to increase with the buckling rate. The average buckling rate of easterly flow is 1.274 that is the smallest among the three, and its duration is also the longest; the average buckling rate of southerly flow is 1.335 that is the largest among the three, and its duration is the shortest. Table 3 The gradient and buckling rate of the main courses of the LYR Index Type Duration Gradient (‰) Buckling rate 1 Northerly flow — 0.316 1.273 Northerly flow 2 612 years 0.344 1.285 Easterly flow 3 1023 years 0.345 1.298 Northerly flow 4 146 years 0.325 1.349 Southerly flow 5 300 years 0.296 1.359 Southerly flow 6 361 years 0.305 1.310 Easterly flow 7 — 0.331 1.250 *The ―—‖ represents that the value of it is unknown or ambiguous. 4 Discussion There are many factors that can influence the occurrence of a course shift, such as gradient, sinuosity, climatic change, human activities and so on. In this study, these factors are classi- fied into two types: internal factors and external factors. 4.1 Internal factors Internal factors are the courses‘ own characteristics, such as sediment rate, gradient, sinuos- ity, topography and so on. This passage mainly discussed the affect of sediment rate, gradi- ent and sinuosity. The fact that middle Yellow River flows through the Loess Plateau brings about heavy sil- tation problem to the LYR. Large amount of silt continuously deposits on the riverbed of the LYR, which results in growing height of riverbed over time. When the height of the riverbed is higher than land, the LYR actually changes to a suspending river. This process may be the main reason that the LYR has frequently breached and changed its flow directions during last thousands of years. Gradient and sinuosity are other factors that may also influence the lifetime of a river course. Figure 10 illustrates that the relationship between the gradient and the lifetime of a course is positive linear, which can be interpreted as the influence of gradient on the flow speed. Qian et al. (1987) has made the same argument that the gradient of a river course is positively correlated with its velocity of flow. A greater flow speed favorably sluices more sediment with the flow into the sea, which reduces the deposit rate (Wang and Yang, 1993b). A low deposit rate can slow down the growing speed of the riverbed of the course and then may extend the river‘s lifetime. As shown in Figure 11, the lifetime of a course is negatively correlated with its sinuosity. The river with higher sinuosity tends to flow more slowly. Thus, a low degree of sinuosity seems helpful to extend its lifetime. Figure 10 The relationship between the gradient Figure 11 The relationship between the buckling and the lifetime of a course rate and the lifetime of a course 4.2 External factors Besides the internal factors, some other external factors also greatly influence the occurrence of course shifts, such as climate change and human activities. 1) Climate change The influence of climate change on the LYR is mainly reflected in precipitation and tem- perature (Wang and Yang, 1993a). This study collected the state of temperature and precipi- tation of 26 course shifts (Table 4) to discuss how the climate change influences course WANG Yingjie et al.: The geo-pattern of course shifts of the Lower Yellow River 1033 shifts of the LYR. As shown in Table 4, 20 out of 26 course shift events occurred in the humid years which probably meant relatively high precipitation in these years. High precipitation tends to pro- duce large runoff and even floods (Liang and Jia, 1991), which creates high risk of breach in the riverbank. On the other hand, large runoff may cause intense riverbed erosion in the up- per and middle Yellow River, which is down washed to produce large amounts of sediment in the LYR (Liang and Jia, 1991), and then increases the possibility of the occurrence of course shifts. Table 4 The comparison between course shifts and climate change Index Climatic period (Temperature) Dry-humid period (Precipitation) Course shift time Diversion point 1 Drought Suxukou Warm 602 B.C. 2 Warm to cold Humid 132 B.C. Huzi Warm to cold Humid Guantao 3 109 B.C. Lingxian Warm to cold Humid 4 39 B.C. Weijun Warm to cold Humid 5 111 A.D. 6 955 A.D. YangguWarm to cold Humid 7 1020 A.D. Warm to cold Humid Huazhou 8 1034 A.D. Warm to cold Humid Tanzhou 9 1048 A.D. Humid Tanzhou Warm to cold 10 1060 A.D. Humid Tanzhou Warm to cold 11 1081 A.D. Humid WeijunWarm to cold 12 1128 A.D. Drought Warm to cold Junxian, Huaxian 13 1168 A.D. DryWarm to cold Ligudu Yangwu 14 Cold Dry 1194 A.D. Yuanwu, Kaifeng 15 1286 A.D. Cold to warm Humid Pukou Caoxian 16 1297 A.D. Humid Cold to warm Yuanwu Kaifeng 17 1344 A.D. Humid Cold to warm Yuanwu, Xingze 18 1391 A.D. Humid Cold to warm Kaifeng Caoxian 19 1416 A.D. Humid Cold to warm Lanzhou 20 1448 A.D. Dry Cold to warm Caoxian 21 1489 A.D. Dry Cold to warm Lanyang Dry 22 1509 A.D. Cold to warm Zhengzhou 23 1534 A.D. Humid Cold to warm 24 1588 A.D. Humid Cold to warm 25 1855 A.D. HumidCold to warm 26 1938 A.D. Warm to cold Humid * ―Warm‖ represents that the year corresponding to the course shift is in a warm period; ―Cold‖ represents in a cold period; ―Warm to cold‖ represents in a transitional period from warm to cold; ―Cold to warm‖ represents in a transitional period from cold to warm. Moreover, 24 out of 26 course shifts happened in the transition periods of cold and warm spells. The climate generally fluctuates more violently in the transitional periods and storms occur more frequently than other periods (Wang and Yang, 1993b). Wang et al. (1993b) also argued that different temperature had influence, to different degrees, on sediment yield, transportation and deposition, and then may affect course shifts of LYR (Wang and Yang, 1993b). 2) Human activities The impact of human activities on the LYR is found mainly in two ways: land use change around the Yellow River Basin (Xu J X, 1993), especially around the middle section of the LYR, and management policy of the Yellow River. a) Xu (2003) arranged the population density of the Loess Plateau from 0 A.D. to 2000 A.D., and concluded that population increased more and more quickly over time. The need for food correspondingly increased with growing population, resulting in more land recla- mation for agriculture. Zhu (1991) studied the historical distribution of agricultural lands on the Loess Plateau based on historical documents, and found that many forests and grassland were destroyed and reclaimed to agricultural areas that intensively expanded over time. De- forestation and expansion of agricultural area promoted soil erosion and then resulted in in- creasing sedimentation in the LYR (Zhu, 1991) that became easier to breach or shift course. b) A usual way of regulating the LYR is constructing high and solid riverbanks to reduce the probability of course shift events. Besides, another famous method to manage the LYR is called ?narrowing the river channel to concentrate water flow and scour sediment‘ put for- ward by Pan Jixun, a famous expert on regulating the LYR in Chinese history. The sedimen- tation rate of the LYR declined significantly when this method was first applied (Xu, 2003), because this method increased flow velocity that made sediment easier to be down washed to the sea. In sum, human activities aim at improving people‘s living standard, avoiding disasters and taking advantage of the Yellow River. Although human activities extend the lifetime of some of the sections of the Yellow River, they also actually add uncertain factors to the changing pattern of the LYR. For example, the course shift in 1048 A.D. running toward north again was mainly interpreted that human blocked the way the river ran toward south. However, these activities only slow down the LYR‘s change, and cannot essentially change the LYR pattern that actually is based on its own rules. Just like the course shift in 1048 A.D., it only lasted 146 years, and then flowed toward south to follow its pattern. 5 Conclusions To study the changing pattern of the LYR, this study arranged the records of the LYR in his- torical documents into digital data for visualization and specific analysis. Based on these data, changing patterns of course shifts of the LYR are summarized as follows. 1) Courses of the LYR are divided into three types based on their flow directions: the northward flow, the eastward flow and the southward flow. 2) The influence scope of the LYR is very large, which is in the shape of sector and al- most equal to the area of France. To the north, it has flowed into the Haihe River, and to the south, it has captured the Huaihe River. Based on the flow types, this area can also be di- vided into three types: the northward flow basin, the eastward flow basin and the southward flow basin. WANG Yingjie et al.: The geo-pattern of course shifts of the Lower Yellow River 1035 3) The shifting velocity of LYR is higher and higher over time, and that of southward flow was much higher than eastward flow and northward flow. 4) The overall changing pattern in flow directions of the LYR is described as follows. It first took its course toward the north by east, then turned to flow to east by north, east by south, south by east in proper order; after it reached the southernmost boundary of the sector, the river tended to flow toward east by south again, and finally turned to flow east by north as the present LYR. Thus, the next course shift of the LYR has a great possibility to shift to north. 5) Most of diversion points are clustered within a triangle with the vertexes named Tao- huayu, Puyang and Tongwaxiang. 6) The relationship between the gradient and the lifetime of a course is positively linear, and that between the sinuosity and the lifetime of a course is negatively linear. 7) The factors that influenced the occurrence of a course shift can be divided into internal factors (sediment rate, gradient, sinuosity and so on) and external factors (climate change, human activities and so on). 8) Most of the course shifts occur in periods with large amounts of precipitation and in the transitional periods of cold and warm spells. However, the exact and quantitative explanations for these changing patterns are still un- clear, which need further study. References Chang J, Liu G H, Liu Q S, 2004. Dynamic monitoring of coastline in the Yellow River Delta by remote sensing. Geo-information Science, 94–98. (in Chinese) Cui B, Li X, Chang X Y et al., 2010. Remote sensing monitoring of coastline change in the Yellow River Estuary (1976–2005). In: Tan Z H, Yi W, Tao X et al. (eds.). 2010 3rd International Congress on Image and Signal Processing (CISP 2010). Yantai, China, IEEE, 2225–2231. (in Chinese) Cui B L, Chang X L, Chen Y L et al., 2007. Dynamic monitoring of coastline in the Yellow River Estuary by remote sensing. Science of Surveying and Mapping, 108–109, 119, 196. (in Chinese) Dai J L, 2005. Encyclopedic Dictionary of Ancient and Modern Chinese Geographical Names. Shanghai: Shanghai Lexicographical Publishing House. (in Chinese) Das J D, Dutta T, Saraf A K, 2007. Remote sensing and GIS application in change detection of the Barak River channel, NE India. Photonirvachak: Journal of the Indian Society of Remote Sensing, 35: 301–312. Downward S R, Gurnell A M, Brookes A, 1994. A methodology for quantifying river channel planform change using GIS. Variability in Stream Erosion and Sediment Transport, 449–456. Editorial Committee of Chinese Physical Geography-Chinese Academy of Sciences (ECCPG-CAS), 1982. The Physical Geography of China. Beijing: Science Press. (in Chinese) Goswami U, Sarma J N, Patgiri A D, 1999. River channel changes of the Subansiri in Assam, India. Geomorphology, 30: 227–244. Gurnell A M, 1997. Channel change on the River Dee meanders, 1946–1992, from the analysis of air photographs. Regulated Rivers: Research & Management, 13: 13–26. Gurnell A M, Downward S R, Jones R, 1994. Channel planform change on the River Dee Meanders, 1876–1992. Regulated Rivers: Research & Management, 9: 187–204. Huang H J, Fan H, 2004. Monitoring changes of nearshore zones in the Huanghe (Yellow River) Delta since 1976. Oceanologia et Limnologia Sinica, 306–314. (in Chinese) Huang W W, Zhang J, Zeng H Z, 1992. Particulate element inventory of the Huanghe (Yellow River): A large, high-turbidity river. Geochimica et Cosmochimica Acta, 56: 3669–3680. (in Chinese) Leys K F, Werritty A, 1999. River channel planform change: Software for historical analysis. Geomorphology, 29: 107–120. Liang J Y, Jia Z F, 1991. A study on the mutual relations among precipitation, runoff and sediment yield on the Loess Plateau. In: Wu X D (ed.). Collection of Researchs on Environmental Changes of the Yellow River Basin and Laws of Water and Sediment Transpotation, Vol.1. (in Chinese) Qian N, Zhang R, Zhou Z, 1987. Fluvial Processes. Beijing: Science Press. (in Chinese) Sarma J N, Borah, D, Goswami U, 2007. Change of river channel and bank erosion of the Burhi Dihing River (Assam), assessed using remote sensing data and GIS. Photonirvachak: Journal of the Indian Society of Remote Sensing, 35: 93–100. Shen Y, Zhao S X, Zheng D L, 1935. The Chronology of Yellow River. Beijing: Military Commission of the Resources Committee. (in Chinese) Wang Y J, Yang P L, 1993a. A preliminary study on the relationship between sloar activity and breaches and course shifts of the lower reaches of Yellow River. In: Wu X D (ed.). Collection of Researchs on Environmental Changes of the Yellow River Basin and Laws of Water and Sediment Transpotation, 4: 62–67. (in Chinese) Wang Y J, Yang P L, 1993b. A study on the relationship between climatic change and evolution characteristics and rules of the Lower Yellow River. In: Wu X D (ed.) Collection of Researchs on Environmental Changes of the Yellow River Basin and Laws of Water and Sediment Transpotation, 4: 6–12. (in Chinese) Wang Y, Yuan K S, Yu Z Y, 2003. Multidimensional Dynamic Visualization of Geo-Information. Beijing: Science Press. (in Chinese) Xu H L, 1993. A study on the changes of water and sediment and the changes of the ancient courses of the Lower Yellow River in historical period. In: Wu X D (ed.) Collection of Researchs on Environmental Changes of the Yellow River Basin and Laws of Water and Sediment Transpotation, 4: 68–76. (in Chinese) Xu J X, 1993. A study of long-term environmental effects of river regulation on the Yellow River of China in historical perspective. Geografiska Annaler Series A: Physical Geography, 75: 61–72. (in Chinese) Xu J X, 2003. Sedimentation rates in the lower Yellow River over the past 2300 years as influenced by human activities and climate change. Hydrological Processes, 17: 3359–3371. Xue Y C, Ma S Y, Zhou C H, 2009. The shoreline migration and the delta changes study in the modern Yellow River Delta using remote sensing and GIS. Marine Sciences, 36–40. (in Chinese) Yang X J, Damen M C J, Zuidam R A V, 1999. Satellite remote sensing and GIS for the analysis of channel migration changes in the active Yellow River Delta, China. International Journal of Applied Earth Observation and Geoinformation, 1: 146–157. (in Chinese) Yao H Y, 1987. A Preliminary Study on the Truth of the Changes of the Yellow River in Last 2700 Years. Xi‘an: Shaanxi Science and Technology Press. (in Chinese) Yellow River Conservancy Commission (YRCC), 1984. Summary History of Yellow River Water Conservancy. Beijing: Water Conservancy and Electric Power Press. (in Chinese) Zhao J M, Li X M, Zhang S Y, 1996. Dynamic monitor the ice flood of Yellow River by remote sensing. Remote Sensing Information, 10, 39–40. (in Chinese) Zhu S, 1991. Environmental changes as influenced by human beings. In: Yang Q (ed.). Natural Environment and its Evolution of the Loess Plateau of China. Beijing: Science Press, 108-143. (in Chinese)