DSpace at VNU: Natural levees and human settlement in the Song Hong (Red River) delta, northern Vietnam

DSpace at VNU: Natural levees and human settlement in the Song Hong (Red River) delta, northern Vietnam tài liệu, giáo á...

Research paper Natural levees and human settlement in the Song Hong (Red River) delta, northern Vietnam The Holocene 22(6) 637­–648 © The Author(s) 2012 Reprints and permission: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0959683611430847 hol.sagepub.com Ayako Funabiki,1 Yoshiki Saito,2 Vu Van Phai,3 Nguyen Hieu3 and Shigeko Haruyama4 Abstract The Song Hong (Red River) delta, northern Vietnam, is characterized by huge natural levees in an area of the delta plain known as the West Floodplain where fluvial sedimentation predominates The natural levees along the Day River, a major distributary of the Song Hong, are comparable in size with those of the main course of the Song Hong The Day River levees are 3–8 km wide and rise 2–5 m above the adjacent backswamps and have played an important role in human settlements since the late Metal age We discussed the relationships among the natural levees of the Day River, delta progradation and the distribution of archaeological sites on the delta plain During the early Holocene, the accumulation of sediment discharged by the Song Hong enhanced both aggradation of the floodplain and river mouth progradation within the drowned valley of the Song Hong Radiocarbon dates from cores, trench exposures, and archaeological sites record a dramatic slowing of aggradation when sea level stabilized during 6–4 cal kyr BP (the Holocene sea-level highstand) As sea level fell to the present level during 4–0 cal kyr BP, the river mouth prograded rapidly toward the Gulf of Bac Bo (Gulf of Tonkin) and the river channels extended seaward In the West Floodplain, lateral accretion overtook vertical accretion to generate the present longitudinal profiles of the Song Hong and Day rivers During this period, human settlements spread across the backswamp and Holocene terrace area, lagging around kyr behind the shoreline migration Keywords accumulation rate, archaeological sites, delta, natural levees, sea-level change, shoreline change Received December 2010; revised manuscript accepted 31 October 2011 Introduction Alluvial floodplains are very sensitive to both global hydroclimatic changes and variability of local factors such as sea-level change, tectonic movement, sedimentation, and human activities Natural levees border almost all river channels that are occasionally flooded Typically, they are wedge-shaped ridges of alluvial sediment that are highest at or near the channel margin and gradually taper downward towards the adjacent backswamps Because they add height to river banks and tend to be densely vegetated, levees help to stabilize channels, and their physical properties are important for engineering structures, human dwellings, and agricultural activity The Song Hong (Red River) delta lies on the western coast of the Gulf of Bac Bo (Tonkin) on the margin of the South China Sea (Figure 1a, b) In terms of area, it is the fourth largest delta in Southeast Asia, after the Mekong, Irrawaddy, and Chao Phraya deltas (Sakurai, 1979; Tsubouchi, 1980) The present Song Hong delta plain has been divided into wave-, tide-, and fluvial-dominated systems on the basis of surface topography and hydraulic processes (Mathers and Zalasiewicz, 1999; Mathers et al., 1996) The fluvial-dominated system is in the northwestern part of the delta and is composed of meandering rivers, natural levees, backswamps, and fluvial terraces The fluvial-dominated area, called the West Floodplain or Casier de Hadong (Gourou, 1936; Sakurai, 1979), contains natural levees along the main Song Hong channel and the Day River, a major distributary (Funabiki et al., 2007) Their natural levees on the delta plain are comparable in size with those of other large rivers in Asia (Brierley et al., 1997; Coleman, 1969; Oketani et al., 2007; Umitsu, 1987; Wu et al., 2005) The natural levees are 3–8 km wide and rise 2–5 m above the adjacent backswamp (Figure 1) The West Floodplain has been a centre of cultivation since the late Metal age (2.5–2 cal kyr BP) when humans began farming on the delta plain (Funabiki et al., 2007; Sakurai, 1979, 1980a, 1980b, 1989) In the alluvial plains on large deltas such as the Huanghe (Yellow River), Ganges-Brahmaputra, and Mekong deltas, natural levees formed after the middle Holocene The Huanghe delta is characterized by huge amounts of sediment, originating from soil erosion in the Loess Plateau in its middle reaches The river bed has been rising continuously as sediment deposits have caused channel avulsions, and now lies higher than the surrounding land in some reaches (Wu et al., 2005) The width of the levees is ~10 km and their height is 5–10 m (Syvitski and Kettner, 2011) In the Ganges-Brahmaputra delta, natural levees were initiated after kyr BP (uncalibrated date), after deltaic sediments filled the subsided Sylhet Basin and formed the present delta plain College of Humanities and Sciences, Nihon University, Japan Geological Survey of Japan, National Institute of Advanced Industrial Science and Technology, Japan Hanoi University of Science, Vietnam National University, Vietnam Graduate School of Bioresources, Mie University, Japan Corresponding author: Ayako Funabiki, College of Humanities and Sciences, College of Humanities and Sciences, Nihon University, 3-25-40, Sakurajosui, Setagaya-ku, Tokyo, 156-8550, Japan Email: bickey@chs.nihon-u.ac.jp Downloaded from hol.sagepub.com at The University of Iowa Libraries on March 19, 2015 638 The Holocene 22(6) Figure 1.  Location map (a) SRTM digital elevation map of the Song Hong delta with isobaths of incised valley during the last glacial maximum (Tanabe et al., 2006) (b) Geomorphic classification map of the Song Hong delta, interpreted from aerial photographs, satellite images, and geological maps (c) Longitudinal profiles of natural levees of the Song Hong and Day rivers The broadest levee, with a width of 20 km and relative height of 2–3 m, was formed by 300 bc Other levees 0.5–2.5 km wide and 2–3 m high were present by the middle of the 18th century, consistent with formation by the shifting of channels (Goodbread et al., 2000; Umitsu, 1987, 1993) In the Mekong delta, natural levees are 0.5–2 km wide and several meters to 10 m high in relation to the neighboring backswamp along both abandoned and active distributaries in Cambodia (Oketani et al., 2007) They were deposited over the last 600–1000 years with accumulation rates of 4.0–6.7 mm/yr This high accumulation rate is thought to reflect natural levee migration and aggradation associated with channel switching or an increase in sediment discharge due to human activities over the past 1000–2000 years (Tamura et al., 2009) The Song Hong delta has a relatively high and wide natural levee system in the West Floodplain (Funabiki et al., 2007), and human settlement has a close relationship with these high levees (Sakurai, 1980a, 1980b) The natural levees in the West Floodplain are thought to have formed after cal kyr BP as a result of sealevel fall and human activities (Funabiki et al., 2007) , like natural levees in other large deltas Here we discuss the age and history of the natural levees based on radiocarbon ages in and around the levees and we consider their relationships to delta evolution, and the distribution of archaeological sites Regional setting Geological setting The Song Hong delta is bordered by mountains composed of Precambrian crystalline rocks and Palaeozoic to Mesozoic sedimentary rocks and lies in a Neogene NW–SE-trending sedimentary basin that is approximately 500 km long and 50–60 km wide (Mathers and Zalasiewicz, 1999; Mathers et al., 1996; Nielsen et al., 1999) The NW–SE alignment of the Red River fault system (Rangin et al., 1995) controls the distribution of the mountainous areas, sedimentary basin, drainage area, and the gross trend of the course of the Song Hong within the delta plain Upstream from the delta plain the sedimentary basin is small, and most sediments flow to the delta plain The basin is filled with Neogene and Quaternary sediments of more than km thickness and is subsiding at 0.04–0.12 mm/yr (Mathers and Zalasiewicz, 1999; Mathers et al., 1996; Tran and Dinh, 2000) The Quaternary sequence unconformably overlies the Neogene deposits and is mostly sand and gravel with subordinate lenses of silt and clay; the sequence thickens seaward to a maximum of 200 m at the coast (Mathers et al., 1996) The Song Hong drowned valley, formed during the last glacial maximum, is mainly to the southwest of the present Song Hong channel (Tanabe et al., 2003a, 2003b) and follows approximately the present course of the Day River (Figure 1a) The narrow, elongate valley is oriented NW–SE; it is approximately 20 km wide and more than 80 m deep at the present river mouth area, and about 20 m deep near Hanoi (Funabiki et al., 2007; Lam and Boyd 2008; Tanabe et al., 2006) The present course of the Song Hong was a distributary of the main channel during the Holocene On the eastern side of the drowned valley there is a buried Pleistocene terrace about 15 m below present sea level (PSL) (Funabiki et al., 2007) The drowned valley is successively filled with lowstand, transgressive, and highstand deposits (Mathers and Zalasiewicz, 1999; Mathers et al., 1996), whereas the surrounding sediments are mostly highstand deposits of massive clay Downloaded from hol.sagepub.com at The University of Iowa Libraries on March 19, 2015 639 Funabiki et al intercalated with peaty organic layers (Lam and Boyd, 2000, 2003) During the last glacial maximum, sea level was about 120 m below PSL It rose to approximately 50, 30, 15, and m below PSL at about 11, 10, 9, and cal kyr BP, respectively (Tanabe et al., 2003b) During 15–9 cal kyr BP, transgressive aggradational and retrogradational fluvial and estuarine sediments were deposited above the coarse-grained lowstand fluvial sediments Delta formation was initiated near Hanoi at about cal kyr BP, coinciding with a deceleration of postglacial sea-level rise (Tanabe et al., 2006) Since then, the deltaic sequences have prograded and expanded laterally to form the present-day Song Hong delta Sea level reached PSL at c cal kyr BP Then, after attaining a high of 2–3 m above PSL during 6–4 cal kyr BP (the Holocene sea-level highstand), sea level fell to PSL over the period 4–0 cal kyr BP (Tanabe et al., 2003b) As the delta prograded, the river-mouth morphology changed from funnel-shaped to a straight coastline at cal kyr BP when the hydrologic regime shifted from a tidedominated bay-head setting to a wave-influenced open-coast setting The rate of progradation then declined from 22 to m/yr and the river mouth became lobate at cal kyr BP, largely in response to an increase of sediment discharge from the Song Hong from 17 Mt/yr at cal kyr BP to 27 Mt/yr at cal kyr BP, and then to 49 Mt/yr from cal kyr BP to the present (Tanabe et al., 2006) The increase of sediment discharge during this period was a result of anthropogenic deforestation of the upper reaches of the Song Hong drainage basin Climate and hydrology The Song Hong drains an area of 160 km2 × 103 km2 (Milliman et al., 1995) It originates in the mountains of Yunnan Province, China, at an elevation above 2000 m (Figure 1a) and flows 1200 km to the South China Sea The total sediment and water discharges from the Song Hong are 100–130 Mt/yr and 120 km3/ yr, respectively (Milliman et al., 1995; Pruszak et al., 2002), and the average sediment concentration in the river water is 1.08 kg/m3 (Tanabe et al., 2006) The water discharge varies seasonally because of the subtropical monsoon climate over most of the drainage area At Hanoi gauging station, water discharge reaches a maximum of about 23,000 m3/s in July–August and a minimum of around 700 m3/s during the January–May dry season Approximately 90% of the annual sediment discharge occurs during the summer monsoon season, at which time sediment concentration in the river water may reach 12 kg/m3 (Mathers and Zalasiewicz, 1999; Mathers et al., 1996) Longitudinal profiles of the Song Hong and Day rivers Longitudinal profiles of two rivers (Figure 1c) were compiled starting from Son Tay, where the Song Hong flows from an area of Pleistocene terraces onto the alluvial plain This figure is based on the elevation of water areas in the river channels as shown in topographic maps, approximating the average water levels About 20 km downstream from Son Tay, the Day River emerges as a distributary of the Song Hong at 10 m above PSL Along the Song Hong in the area of elevated levees, which occupies the northern half of the area between the Song Hong and Day rivers, the average gradient is 1/10,000 and it falls from 10 to m elevation The average gradient of the Song Hong across the West Floodplain is 1/100,500 The profile of the Day River in the area of elevated levees is similar to that of the Song Hong, and the average channel gradient is 1/12,500 with elevation declining from 10 to m From the point of divergence of the two rivers to Phu Ly at the southern limit of the West Floodplain, the average gradient of the Day River, at 1/125,000, is gentler than that of the Song Hong During 1934–1937, a movable dam and sluice gates were constructed to isolate the Day River from the Song Hong (Bureau of Flood Control Economic Commission for Asia and the Far East, 1951; Sakurai, 1979) The dam is usually closed so that the water source of the Day River is limited to the West Floodplain During severe flooding in 1971, the dam was opened to divert some of the flow of the Song Hong and save Hanoi City from severe flood damage (Haruyama et al., 2003; Sakurai, 1979) The water level of the Day River during the 1971 flood (Figure 1c) indicates that the river carried more water then than now In historical documents, the Day River is often referred to as a main shipping route from Hanoi to the South China Sea (Sakurai, 1980b) Method We used aerial photographs, satellite images, and geological maps to define the areal extent of natural levees, backswamps, and abandoned channels In Figure 2, the red line indicates limit of main natural levees, which are elevated comparing the adjacent backswamps From the interpretation of aerial photographs and satellite images, we illustrated the abandoned channel courses They are located on both natural levees and backswamps forming the meandering belts Using GIS software TNTmips (MicroImage, Inc.), we plotted the locations of archaeological sites (from the data base of Nishimura and Nishino, 2003), previously studied cores (PD and DA; Funabiki et al., 2007), and trench exposures (HN-1 and CS-LAM; Lam and Boyd, 2008) on a Shuttle Radar Topography Mission (SRTM) digital elevation map of the study area (Figure 2) Although the width of natural levees of both Song Hong and Day River are almost same in the West Floodplain (Figure 3), the incised valley course (Figure 1a) and longitudinal profiles (Figure 1c) suggest that the Day River should be the former main course of the delta Thus, we chose the location in the West Floodplain mainly along the Day River, for collection of additional core and trench exposure samples for determination of radiocarbon ages These data were collected and analyzed during 2003–2006 During core and trench exposure sampling we compiled detailed logs of lithology and sedimentary structures, which we compared with the results of previous studies (Funabiki et al., 2007; Hori et al., 2004; Jusseret et al., 2009; Lam, 2003; Tanabe et al., 2003a, 2003b, 2006) Then, we combined the sedimentary facies and radiocarbon dates from our samples and the previous studies mentioned above (Figures and 3) We obtained 16 new radiocarbon ages (analyses by Beta Analytic Inc.) from peaty organic material, freshwater shells, and organic sediments in the cores we collected From these and the data from cores PD and DA (Funabiki et al., 2007), we determined calibrated radiocarbon ages by using the CALIB 6.0.1 algorithm of Stuiver et al (1998) Some samples were too small to determine δ13C However, all conventional radiocarbon ages were corrected for isotopic fractionation effects All ages are reported as calibrated radiocarbon ages (cal kyr BP) unless otherwise noted We also compiled radiocarbon ages determined in previous archaeological studies (Nishimura and Nishino, 2003) and those determined from trench exposures by Lam and Boyd (2008) They are reported as measured ages and are from peat, wood, and charcoal For conventional radiocarbon ages, we used δ13C of −25±2‰ for wood, organic material, and charcoal and −27±3‰ for peat and organic sediment following other measured data in this area as well as Polach (1976) and Stuiver and Polach (1977) The conventional radiocarbon age was calculated using the radiocarbon decay equation (Stuiver and Polach, 1977) A spreadsheet with the equation for this calculation was taken from the Calib website at http://calib.qub.ac.uk/calib/manual/chapter5 html Calibrated radiocarbon ages were calculated from the conventional age using CALIB 6.0.1 Downloaded from hol.sagepub.com at The University of Iowa Libraries on March 19, 2015 640 The Holocene 22(6) Figure 2.  SRTM digital elevation model of the West Floodplain showing locations of trench exposures, cores, archaeological sites used in this study, and cross sections shown in Figure To compare radiocarbon ages from archaeological sites to those obtained from trench exposures and cores, we needed to reconsider the prehistorical chronology defined by Nishimura and Nishino (2003), who described the prehistorical periods of the Song Hong delta as (1) Early Neolithic age of the Da But culture (6.5–4.5 kyr BP); (2) Late Neolithic age of the Phung Nguyen culture (4.5–4 kyr BP); (3) Early Metal age of the Dong Dau (3.1–2.8 kyr BP) and Go Mun (2.7–2.4 kyr BP) cultures; and (4) Late Metal age of the Dong Son culture (2.5–2 kyr BP) These divisions were all based on uncalibrated ages from wood or charcoal at archaeological sites Considering these divisions and our calibration of the archaeological radiocarbon data of Nishimura and Nishino (2003) (see Appendix), we defined the early Neolithic, late Neolithic, early Metal, and late Metal ages as early Holocene–4.5 cal kyr BP, 4.5–3.5 cal kyr BP, 3.5–2.5 cal kyr BP, and 2.5–2 cal kyr BP, respectively We set the beginning of the early Neolithic in the early Holocene, because the Hoabininian culture, which preceded the early Neolithic age, ended 10–7 kyr BP (uncalibrated) (Nishimura and Nishino, 2003) In our discussion, we use these ages for those archaeological sites that lack accurate radiocarbon ages Based on the lithological data from cores and trench exposures, radiocarbon ages, and archaeological data, we compiled cross-sections showing approximate isochrons (Figures and 3) Four cross-sections are oriented SW–NE, perpendicular to the gross course direction of the Song Hong and Day rivers and their natural levees, and three are oriented NW–SE, subparallel to the long axes of the levees Accumulation rates were calculated from the radiocarbon data to examine the rate of formation of the natural levees (see Figure 5) From the data described above, we considered the process of formation of the natural levees in relation to shoreline change and the distribution of archaeological sites (see Figures and 6) Results Below, we referred seven cross-sections of the West Floodplain Each section indicates the present- and paleo-morphology of the natural levees and baskswamp areas Downloaded from hol.sagepub.com at The University of Iowa Libraries on March 19, 2015 641 Funabiki et al Figure 3.  Cross-sections of the West Floodplain HT, HN and VP indicate Ha Tay, Hanoi, and Vinh Phuc provinces, respectively For example, HT40 indicates archaeological site no 40 in Ha Tay Province, as described by Nishimura and Nishino (2003) (a) Transverse cross-sections (b) Longitudinal cross-sections Locations of cross-sections are shown in Figure Because the SRTM digital elevation model accentuates the undulation, we used the elevation of topographic map Downloaded from hol.sagepub.com at The University of Iowa Libraries on March 19, 2015 642 The Holocene 22(6) Line A Line A runs through the natural levees of the Day River and Song Hong, which have oxbow lakes and abandoned channels on them This line crosses the widest part of the Day River natural levees (about km) Both the southwestern and northeastern ends of the line terminate on the Pleistocene terraces at an elevation above 10 m On the southwestern Pleistocene terrace, archaeological site HT29 is located at 8.75 m in elevation with a cultural layers of late Metal age (2.5–2 cal kyr BP, depth unknown) Site HT25 shows evidence of late Neolithic (4.5–3.5 cal kyr BP) settlement at 8.6 m in elevation on the natural levees of the Song Hong On the northeastern part of line A, archaeological site VP28 provides a late Metal age (2.5–2 cal kyr BP, 10.1–10.2 m above PSL) At site HN3, a late Neolithic settlement was found at 9.0 m in elevation They are located on the northeastern Pleistocene terrace, which is incised by the Ca Lo River Line B Line B is located 15 km south of the point of divergence of the Song Hong and Day rivers It starts at the Pleistocene terrace and crosses the natural levees of the Day River, adjacent backswamp, and natural levees of the Song Hong This line ends at the Pleistocene terrace on the left side of the Song Hong The sediment thickness after the middle Holocene show less than m in the Pleistocene terrace area, while it is m at maximum in the backswamp area Trench exposure HN extends to a depth of 1.1 m We drilled core HN to a depth of 0.9 m beyond the bottom of the trench exposure The core consists of mottled reddish brown clay that has been interpreted as a backswamp deposit (Fsc of Miall, 1992) or floodplain deposit with lateritic soils (Collinson, 1996; Funabiki et al., 2007; Tanabe et al., 2006) We obtained radiocarbon ages of 7.2 and 0.7 cal kyr BP at elevations of 5.7 and 6.4 m, respectively An earthenware fragment from the Ly Dynasty (1010–1225) was found in the trench exposure at an elevation of 6.1 m Archaeological sites HT42 and HT43 contain cultural layers from the late Metal (2.5–2 cal kyr BP) and late Neolithic (4.5–3.5 cal kyr BP) age, respectively, both at elevations higher than m, indicating that less than m of sediment has been deposited since the middle Holocene HT43 site is covered with yellow soil and gravels (Nishimura and Nishino, 2003) indicating the modern flood sediments from the surrounding Pleistocene terrace area In the backswamp, trench exposure LY is composed of a black peaty layer overlain by greyish blue to red mottled clay with abundant rootlets and gypsum crystallization at the top It has been interpreted as floodplain deposits with lateritic soils (Collinson, 1996, Tanabe et al., 2006) We obtained a radiocarbon age of 5.8 cal kyr BP from the peaty layer at m above PSL Archaeological site HT40 contains cultural layers of early Metal age (3.5–2.5 cal kyr BP, 3.2 m above PSL) and late Metal age (2.5–2 cal kyr BP) Charcoal from 4.4 m above PSL gave an age of 2.9 cal kyr BP In the area of natural levees and abandoned channels, archaeological site HN53 provided a late Neolithic age (4.5–3.5 cal kyr BP, 7.0 m above PSL), and at site HT26, a late-Neolithic stone artifact was found but its original depth is unknown Line C Line C extends northeast from a limestone mountain that marks the western edge of the West Floodplain The area between the mountains and the Day River is marshy and is often inundated during the rainy season This line goes through the Ho Tay lake, the widest part of natural levees of Song Hong (about km wide), and adjacent backswamp Core TP is composed of reddish brown silt and sand with rootlets and gypsum crystallization at the top, and is intercalated with well-sorted very fine to fine sand beds at the bottom Because of poor core recovery, sedimentary structures could not be identified Sand beds in levee sediments, such as these, have been interpreted as possibly representing flood events (Miall, 1996) or crevasse splay sediments (Tanabe et al., 2006) Sediment at the top of this site is interpreted as backswamp deposits (Fsc of Miall, 1992), or floodplain deposits with lateritic soils (Collinson, 1996; Funabiki et al., 2007; Tanabe et al., 2006) We obtained radiocarbon ages of 9.8, 6.6, 5.7, and 5.3 cal kyr BP from organic sediment at elevations of 3.7, 5.3, 5.7, and 6.9 m, respectively The relatively old radiocarbon age of the highest sample suggests that, because this area is very close to the Pleistocene terrace, the Holocene basement might be shallower and the natural levee sediments thinner than in other areas At site TH, radiocarbon ages were obtained of 6.8 cal kyr BP from a greyish blue silty clay at an elevation of 0.5 m, and 7.4 cal kyr BP from a reddish brown mottled silty sand at 2.5 m This sequence has been interpreted as floodplain deposits with lateritic soils (Collinson, 1996; Tanabe et al., 2006) The lower part of trench exposure TV consists of an upwardfining succession from fine sand layers with climbing ripples to silt to clay, which indicate a high-sediment supply and relatively high-energy environment at the lower part The succession might be interpreted as cut-off meandering channel fill sediments of a Song Hong distributary (Tanabe et al., 2006), or crevasse splay sediments We obtained a radiocarbon age of 7.4 cal kyr BP from organic silty sediment intercalated with the sand layers at 1.4 m elevation Radiocarbon ages of 6.7 and 6.6 cal kyr BP were obtained from wood and organic clay at 2.3 m elevation at site TV The radiocarbon age of the wood is 0.1 kyr older than that of the organic clay Two Holocene aquifer channels have been recognized east of the present Nhue River (Jusseret et al., 2009) They consist of gravel to fine sand and mottled reddish brown clay and are separated by clayey floodplain sediments A series of abandoned channel courses beside the TV site and along the modern distributary (Figure 2) indicate that these sites are located on the filled cut-off meander channel of Song Hong and its distributaries At archaeological site HN55, burial relics of late Metal age (2.5–2 cal kyr BP, depth unknown) were found Between the natural levees of the Song Hong and Pleistocene terrace area, backswamp lies at the elevation of 5–7 m with archaeological sites, called Co Loa Archaeological sites group (Nishimura and Nishino, 2003) Site HN35 is located on the eastern edge of natural levees at the elevation of m and shows a cultural layer of late Neolithic age (4.5–3.0 cal kyr BP, depth unknown) At site HN6, cultural layers of late Metal age (2.5–2 cal kyr BP) located at the elevation of 4.5–4.7 m The sites HN25, 22, were found at the elevation of 5–7 m and contain the late Metal age cultural layers (2.5–2 cal kyr BP, depth unknown) At site HN14, cultural layers from late Neolithic to late Metal were found at 5.6–6.5 m in elevation Line D On line D, the natural levees of the Day River and Song Hong have elevations of 7–9 m, which is 2–4 m higher than the adjacent backswamp This line runs from the natural levees of the Day River, continues through the backswamp and natural levees of Song Hong, and ends at the adjacent backswamp Trench exposure sites BG and DM both show repetitive sequences of fine sand, silt, and clay, interlaminated with peaty layers Bioturbation or roots have disturbed the laminations in some places The sequence is typical of the levee sediments classified as facies code Fl by Miall (1996) Each sedimentary Downloaded from hol.sagepub.com at The University of Iowa Libraries on March 19, 2015 643 Funabiki et al greyish blue clay overlies the peat layer Late Neolithic (4.5–3.5 cal kyr BP) cultural layers were identified at elevations of 4.7 m at site HN65 and m at site HN60, but their original depth is unknown The backswamp area on the left side of the Song Hong are located at the elevation of 4–5 m, relatively lower than the right side The archaeological sites HN46 and 47 show the late Metal age cultural layers (2.5–2 cal kyr BP, depth unknown) Line E Line E starts from the Pleistocene terrace at the northern edge of the Song Hong Delta plain This line crosses the natural levees of the Song Hong and Day River, and backswamp between two rivers It ends at the backswamp on the left side of the Song Hong On the Pleistocene terrace, late Metal age (2.5–2 cal kyr BP) and late Neolithic to early Metal (4.5–2.5 cal kyr BP) cultural layers were found at archaeological sites VP50 and 49, but depth is unknown At site VP54, brick tomb of late Metal (2.5–2 cal kyr BP) was found at 9.6–9.5 m in elevation The natural levees of the Song Hong are about km From the divergence point of two rivers to southward, the natural levees were fed by deposition of both two rivers Archaeological sites HT25, HT26, and HN53 show evidence of late Neolithic (4.5–3.5 cal kyr BP) settlement at 8.6 m, unknown, and 7.0 m elevations, respectively Along the southern part of line E, Holocene channel-fill sediments lie beneath the natural levees of the Song Hong at site TV Sediments younger than cal kyr BP are grey to reddish mottled silty clay with black peaty layers and abundant rootlets, and have been interpreted as backswamp deposits (Fsc of Miall, 1992), or floodplain deposits with lateritic soils (Collinson, 1996; Funabiki et al., 2007; Tanabe et al., 2006).There are no archaeological sites on the natural levees of the Song Hong and backswamps on the left side of the Song Hong Line F Figure 4.  Selected photographs of natural levee sediments at the trench exposures BG site (5.77–5.12 m in elevation) and a general view of site BG The sediments show repetitive alternation of fine sand and silt to clay S: fine sand layer Sand-mud couplets represents a flood event cycle represents a flood event They reflect riding-and waning-stage deposits (e.g Iseya and Ikeda, 1989) (Figure 4) At site DM, radiocarbon ages of 5.7 and 3.3 cal kyr BP were obtained from peaty layers at elevations of 4.0 and 1.8 m, respectively At site BG, radiocarbon ages of 5.5, 3.9, and 3.4 cal kyr BP were obtained from organic sediments at elevations of 4.4, 5.4, and 6.1 m, respectively At 4.4 and 6.1 m, freshwater shell fragments (Viviparidae gen et sp indet.) indicate an age of about 0.2 cal kyr BP Because these ages are much younger than those of the organic sediments, we concluded that the sediments had been reworked At archaeological sites HT54, 55, and 56, a cultural layer of late Metal age (2.5–2 cal kyr BP) was found (depth unknown) At site HT57, a wood fragment from a late Metal age cultural layer at m elevation gave a radiocarbon age of cal kyr BP At trench exposure CS-LAM (Lam and Boyd, 2008), a radiocarbon age of 7.4 cal kyr BP was obtained from peat at 2.3 m elevation A 2.7 m thick Line F extends southeast from the Pleistocene terrace After crossing the Day River natural levee, this line runs longitudinally through the approximate centre of the backswamp region between the Song Hong and Day rivers, the natural levees of the Song Hong, and ends at the adjacent backswamp on the left side of the Song Hong Late Metal age (2.5–2 cal kyr BP) cultural layers were found at elevations of 5.5, 1.8, and 4.0 m at sites HT34, HT31, and HT39, respectively At site HT31, sandy soil covered the cultural layers (Nishimura and Nishino, 2003), indicating relatively high energy environment This sandy soil might be influenced by the modern flood sediments from the Day River and distributaries At sites HT33 and HT40, cultural layers of late Neolithic age (4.5– 3.5 cal kyr BP) were found at elevations of 5.1 and 3.2 m, respectively At site HT40, a radiocarbon age of 2.9 cal kyr BP was obtained from charcoal at an elevation of 4.4 m On the southern part of line F, late Metal age (2.5–2 cal kyr BP) settlements (HT54, HT56, and HT57) can be observed and site HT is shallower than m depth In core DA, sediment younger than 6.3 cal kyr BP forms a sequence of bioturbated grey to reddish brown clay, which is up to 3.5 m thick and contains peaty organic layers in its lower part and abundant rootlets in its upper part This sequence corresponds to a lateritic weathering profile developed in floodplain and channel-levee sediments at the land surface of the core site and has been interpreted as modern floodplain sediments (Funabiki et al., 2007) and overlies tide-influenced channel-fill sediments of a deltaic sequence A radiocarbon date of 6.3 cal kyr BP was obtained from floodplain sediments from core DA (Funabiki et al., 2007) at m above PSL Downloaded from hol.sagepub.com at The University of Iowa Libraries on March 19, 2015 644 The Holocene 22(6) Line G not take into account sediment compaction On the basis of these figures, we first discuss sediment accumulation during the early Holocene (before cal kyr BP) and then discuss natural levee formation since the middle Holocene during a time of stable sea level Lastly we discuss the relationship between natural developments and human settlement Line G runs from the Pleistocene terrace that marks northwestern edge of the West Floodplain This line crosses the Day River for several times at the elevation of 7.5–10 m and contains natural levee sediments On the southern part of this line, the center of the West Floodplain lies below m in elevation The southeastern end of this line reaches the natural levees of the Song Hong and adjacent backswamp At archaeological site HT17, the earliest evidence of a late Metal age settlement (2.5–2 cal kyr BP) is below m elevation In core PD, mangrove swamp sediments older than cal kyr BP lie below PSL and are overlain by m of reddish brown floodplain sediments (Funabiki et al., 2007) However, the adjacent natural levee sediments gave radiocarbon ages younger than cal kyr BP at m elevation Thus, we inferred that there was a local surface depression in the vicinity of the PD site at cal kyr BP The earliest evidence of early Metal age settlements (3.5–2.5 cal kyr BP) at site HT42 is at an elevation of 5.5 m Fine reddish brown clay covered this site (Nishimura and Nishino, 2003), which is interpreted as floodplain sediments or surface soil A miniature bronze drum (late Metal age, 2.5–2 cal kyr BP) was found at site HT51 on a secondary natural levee at m elevation, but the depth from which the drum was excavated is unknown Bronze drums (late Metal age, 2.5–2 cal kyr BP) were also found m above PSL on backswamp at sites HT58 and HT98 The excavation depths of the drums are unknown Discussion From the radiocarbon dates of the trench exposure and data from archaeological sites, we derived isochrons at the millennial scale (Figure 3) and sediment-accumulation curves (Figure 5), but did Sediment accumulation during the early Holocene Figure shows sediment accumulation and sea-level change during the Holocene Early Holocene sea-level rise continued until approximately cal kyr BP At cal kyr BP, the sea level was m below PSL and rising, and at cal kyr BP it reached 2–3 m above PSL (Tanabe et al., 2003b) Before this highstand, accumulation of estuarine and deltaic sediments was very rapid, with rates of 1.9–31.9 mm/yr (average mm/yr), and kept pace with sea-level rise This high accumulation rate indicates rapid aggradation of estuarine to deltaic sediments into a drowned valley (Hori et al., 2004; Tanabe et al., 2006) Similar aggradation during Holocene delta initiation is reported from the Mekong delta (Tamura et al., 2009) Floodplain accumulation since the middle Holocene Before discussing sediment accumulation in the West Floodplain, we will summarize sediment accumulation based on radiocarbon dates and archaeological data for each profile line in Figure The archaeological site HT25 in line A indicates low accumulation for the last 3–4 kyr on the natural levees of Song Hong In the western part of line B, the Pleistocene terrace area, sites HN, HT42 and 43 show that there were very thin sediments from the last cal kyr The accumulation rate appears to havebeen high Figure 5.  Sediment-accumulation curves derived from radiocarbon ages obtained from cores, trench exposures, and archaeological sites The sea-level curve is modified from that of Tanabe et al (2006) Data sources and coordinates of sample locations are provided in Table Downloaded from hol.sagepub.com at The University of Iowa Libraries on March 19, 2015 645 Funabiki et al Table 1.  Summary of radiocarbon ages obtaimed from sediment cores, trench exposures, and archeological sites on the West Floodplain CS-LAM and HN-1 trench exposure ages were from Lam and Boyd (2008) Cores DA and PD ages were from Funabiki et al (2007) Archaeological site ages were from the data base of Nishimura and Nishino (2003) Site ID  Outcrops TV     TH   BG   Lat (N) Long (E) 20°59′35″ 105°47′22″ 6.8 4.50 4.50 5.40 2.50 4.50 1.40 1.40 2.30 2.30 1.40 2.50 0.50 6.10 6.10 2.10 3.10 5.40 4.40 3.10 0.80 1.80 4.00 2.10 2.70 4.00 2.80 1.90 4.40 6.70 5.70 3.50 4.00 3.40 1.00 2.25 3.00 0.80 2.00 2.40 1.10 1.80 0.77 1.88 3.50 6.20 9.53 12.60 13.75 15.20 4.70 8.50 10.10 10.92 11.70 13.20 14.09 15.51 6.90 5.70 5.30 6.40 5.70 4.73 3.62 2.00 -0.70 -4.03 -7.10 -8.25 -9.70 2.80 -1.00 -2.60 -3.42 -4.20 -5.70 -6.59 -8.01 1.00 4.00 1.00 4.00 0.65 4.35 1.80 3.20 HT36 (Chua 21°03′40″ 105°43′20″ 5.9 Gio) 1.70 4.20 HN63 (Van Dien) 1.65 3.35 21°00′31″ 105°46′32″ 20°55′21″ 105°43′30″ 7.5       DM 20°55′21″     AK 21°00′14″   LY 21°01′39″ CS-LAM 20°57′29″ HN-1 21°01′18″ Cores TP 20°57′17″     HN 21°00′08″   DA 20°54′08″               PD 21°01′25″               Archeological sites HT57 (Phu 20°56′39″ Luong)   HT40 (Go Chien Vay)   Conventional δ13C age (‰) (yr BP) Calibrated age (2σ range) (cal yr BP) Reference organic sediment wood organic sediment organic sediment organic sediment organic sediment shell fragment Viviparidae gen et sp indet organic sediment shell fragment Viviparidae gen et sp indet organic sediment organic sediment organic sediment organic sediment organic sediment organic sediment charred material peat peat 6040 ± 50 5860 ± 50 6440 ± 50 6440 ± 50 5950 ± 50 4820 ± 40 240 ± 40 -28.2 -27.6 -28.6 -27.4 -29.1 -21.9 -5.9 6660 (6370-6950) 6685 (6630-6740) 7360 (7320-7400) 7360 (7320-7400) 6750 (6700-6800) 5540 (5480-5600) 160 (0-320 ) this study this study this study this study this study this study this study 3600 ± 40 180 ± 40 -23.1 -5.7 3915 (3860-3970) this study 110 (0-220 ) this study 3160 ± 40 3060 ± 40 5010 ± 40 8760 ± 50 5630 ± 50 5610 ± 60 5101 ± 40 6470 ± 130 6200 ± 130 -22.3 -23.4 -22.4 -21.8 -22.9 -23 * -27.5 -27.5 3390 (3360-3410) 3300 (3240-3350) 5770 (5660-5870) 9790 (9680-9890) 6390 (6320-6460) 6380 (6320-6440) 5840 (5760-5910) 7370 (7270-7480) 7100 (6950-7260) this study this study this study this study this study this study this study Lam (2003) Lam (2003) organic sediment organic sediment organic sediment charred material organic sediment organic material organic material organic material organic material organic material organic material organic material organic material organic material wood wood organic material organic material organic material organic material organic material 4760 ± 40 4990 ± 40 6190 ± 40 810 ± 40 6270 ± 40 3810 ± 40 3290 ± 40 5550 ± 70 6490 ± 50 6350 ± 50 7200 ± 50 7840 ± 60 7920 ± 60 7450 ± 60 4560 ± 50 6040 ± 50 6390 ± 50 6600 ± 50 6700 ± 50 6650 ± 50 7200 ± 50 -21.4 -22.1 -24.8 -27.1 -24.7 -17.73 -22.26 -26.94 -28.61 -31.52 -28.53 -28.37 -28.38 -21.99 -28.74 -29.17 -26.89 -28.2 -28.41 -28.26 -27.56 5520 (5470-5560) 5750 (5660-5840) 7090 (7020-7160) 720 (680-760) 7210 (7170-7250) 4210 (4150-4260) 3520 (3470-3570) 6380 (6360-6400) 7330 (7250-7400) 7290 (7250-7320) 8000 (7960-8040) 8600 (8540-8660) 8710 (8630-8780) 8300 (8260-8330) 5150 (5120-5180) 6900 (6850-6950) 7300 (7270-7330) 7480 (7460-7500) 7590 (7560-7610) 7560 (7500-7610) 8010 (7960-8050) this study this study this study this study this study Funabiki et al., (2007) Funabiki et al., (2007) Funabiki et al., (2007) Funabiki et al., (2007) Funabiki et al., (2007) Funabiki et al., (2007) Funabiki et al., (2007) Funabiki et al., (2007) Funabiki et al., (2007) Funabiki et al., (2007) Funabiki et al., (2007) Funabiki et al., (2007) Funabiki et al., (2007) Funabiki et al., (2007) Funabiki et al., (2007) Funabiki et al., (2007) wood (late Metal age cultural layer) wood (late Metal age cultural layer) charcoal (early Metal age cultural layer) settlement/ burial (early Metal age cultural layer) settlement (early to late Metal age cultural layer) settlement (late Neolithic age cultural layer) 2020 ± 120 -25 2110 ± 140 -25 2310 ± 210 -25 2000 (1860-2130) Nishimura and Nishino (2003) 2070 (1930-2210) Nishimura and Nishino (2003) 2360 (2100-2620) Nishimura and Nishino (2003) Nishimura and Nishino (2003) Nishimura and Nishino (2003) Altitude Depth Altitude Material (m) of site (m) (m) 105°43′30″ 7.5 105°42′33″ 6.1 105°42′56″ 105°48′04″ 105°46′04″ 4.9 105°42′13″ 7.7 105°37′49″ 7.5 105°49′29″ 5.5 105°38′25″ 7.5 105°46′16″ 21°02′00″ 105°43′00″ 20°56′40″ 105°49′40″ c 3500-2300 c 4500-2000 c 4500-3500 * δ13C not calculated because the sample was too small Downloaded from hol.sagepub.com at The University of Iowa Libraries on March 19, 2015 Nishimura and Nishino (2003) 646 The Holocene 22(6) for the last kyr with ~1 m of sediment, which consists of mostly surface soil The backswamp area in the middle part shows relatively high accumulation rates for the last kyr in low-lying areas However, the eastern part of the backswamp indicates no deposition for the last 3–4 kyr, which includes natural levees along the Song Hong (sites HN53 and HT26) with relatively high elevation In line C, the accumulation curve of site TP at the edge of natural levees of the Day River indicates very rapid aggradation before cal kyr BP, followed by very little or no accumulation for the last cal kyr The backswamp area in the middle part shows relatively slow accumulation, less than m/kyr In line D, although site BG near the Day River shows high accumulation rates at 1–0 kyr BP, this site is located in modern dykes and near main channels, thus it is interpreted as modern processes and reworking The backswamp area in the middle part of line D shows ages of 2–4 cal kyr BP for surficial sediments (e.g sites HT57 and HN65), indicating that sedimentation was active before 2.5–2 cal kyr BP and not active for the last kyr In line E, the northern part (Song Hong natural levees) and the middle backswamp area indicate that aggradation was active before ~4 kyr BP Line F shows more or less continuous accumulation until kyr BP and intermittent accumulation to the present If the age of the surface of at site DA (elevation 5.5 m) is present, the site shows almost constant and continuous accumulation since 6.5 kyr BP Line G shows a variety of accumulation patterns Natural levee areas of the Day River show active accumulation before kyr BP at site TP and 1–0 kyr BP at site BG, and active accumulation before kyr BP for the southern part After kyr BP, high accumulation rates of more than m/ kyr (1 mm/yr) are only recognized at sites TP (5–6 kyr BP), HT31 (2–0 kyr BP), BG (1–0 kyr BP), and only for natural levees or in-channel sediments All other sites from sedimentary environments of natural levees, floodplains and backswamps show less than m/kyr accumulation The contrast in accumulation rates before and after cal kyr BP demarcates two clear stages in the history of the delta In Figure 2, abandoned channels are located on both of the natural levees and backswamp areas forming the meandering belts They suggest lateral accretion of the Day River and Song Hong after the middle Holocene In general, rivers aggrade in response to relative sea-level rise and degrade in response to relative sea-level fall (Mackin, 1948; Petter and Muto, 2008) Mackin (1948) divided the causes of aggradation of the river channel into ‘upriver or watershed control’ and ‘downriver or base-level control’ In the incised valley of the Song Hong before the sea-level highstand at cal kyr BP, the rise in sea level caused aggradation by ‘base-level control’ and created ample accommodation space in fluvial to coastal plains However, during stable sea level at 6–4 kyr BP and slightly falling sea level for the last kyr, accommodation space was very limited; moreover, sea-level fall induced emergence of Holocene marine terraces (Tanabe et al., 2003b) Abandonment of the Day River natural levees at site TP site is linked to middle Holocene sea-level fall The longitudinal profiles of the Day and Song Hong rivers show clear differences in water level (Figure 1c) The current Song Hong flows at a low level in relation to middle Holocene marine terraces, evidence of slight incision to fit the present sea level However, the Day River shows a higher water level, representing the abandoned natural levee level, therefore it may be occupying a relict geomorphology representing a higher middle Holocene sea level According to Tanabe et al (2006), ‘watershed control’ has occurred mainly after cal kyr BP, increasing the sediment discharge from 17–27 to 49 million t/yr because of deforestration and human activity (Li et al., 2006) As increased accumulation rates for the last kyr are recognized only around sites LY and HT31, other than these local flooding sites fluvial sediments were directly transported to the sea resulting in delta development Shoreline change and distribution of archaeological sites The interpreted palaeogeography of the Song Hong Delta at 8, 4, and cal kyr BP and the corresponding distributions of archaeological sites (Nishimura and Nishino, 2003) are shown in Figure Delta formation was initiated near Hanoi at about cal kyr BP, coinciding with a deceleration of postglacial sea-level rise (Hori et al., 2004; Tanabe et al., 2006) Since then, the deltaic sequences have prograded and expanded laterally to form the present-day Song Hong delta At cal kyr BP, the delta was funnel-shaped (Figure 6a) and most of the sediment discharged by the Song Hong began to accumulate in the river mouth, causing seaward progradation within the Song Hong drowned valley (Tanabe et al., 2006) The shoreline was near the site of core DA in the West Floodplain (Funabiki et al., 2007) Archaeological assemblages found in the mountains and on Pleistocene terraces at elevations above 10 m indicate that Hoabinhian (about 30 kyr BP to early Holocene) and preNeolithic (Late Pleistocene to early Holocene) settlements existed there at cal kyr BP (Figure 6a) At the cessation of sea-level rise (6 cal kyr BP), the river mouth morphology changed from funnel-shaped to a relatively straight coastline because the hydrologic regime shifted from a tide-dominated bay-head setting to a wave-influenced open-coast setting The rate of progradation had decreased from 22 to m/yr by this time (Tanabe et al., 2006), which corresponds to the end of aggradation of floodplain sediments including natural levees (Figure 5) After cal kyr BP, settlements of late Neolithic age (4.5–3.5 cal kyr BP) were widespread on the West Floodplain, where their elevations above the southern coastal areas made them a suitable human habitat In the West Floodplain, nine archaeological sites were located on the edge of natural levees, but not between the present artificial dykes, where is most elevated comparing the adjacent backswamp In backswamp area, there are only three archaeological sites of the late Neolithic age No archaeological sites can be seen on the eastern side of the present Song Hong The distribution of archaeological sites indicates that emergence of the western delta plain preceded that of the eastern delta plain This occurred because the fluvial sediment flux in the west was greater than in the east and the earlier development of natural levees in the west made the floodplain more suitable for human settlement (Funabiki et al., 2007) There are also some archaeological sites on beach ridges, where progradation of deltaic sediments began to develop a lobate delta in a wave-dominated open-coast setting (Figure 6b) By cal kyr BP, land cultivation had commenced at human settlements, which spread across the backswamp area of the West Floodplain and Holocene terraces in the northeastern part of the delta plain (Figure 6c) In the West Floodplain, archaeological sites spread mainly on the backswamp area Eight of the late Metal age archaeological sites were found on the natural levees and 23 sites were recognized on the backswamp area In the coastal areas, settlement sites spread into these lowlands after the shoreline prograded seaward and the natural levees formed by cal kyr BP Thus, there is a lag of kyr between the settlement spread and shoreline migration This time gap may be attributed to the physical conditions of the newly formed delta plain, which were not optimal for settlement because of wetlands, saltwater intrusion, and unstable shoreline The Song Hong delta plain was developed earlier than those of other large deltas in southeast Asia, as early as cal kyr BP Downloaded from hol.sagepub.com at The University of Iowa Libraries on March 19, 2015 647 Funabiki et al Figure 6.  Palaeogeographic maps at 8, 4, and cal kyr BP showing the distribution of archaeological sites (Nishimura and Nishino, 2003) Modified from Funabiki et al (2007) to include new data Palaeo-water depths are relative to the sea level of the time A sea-level curve modifid from Tanabe et al (2006) and the time ranges of archaeological data defined in this study are also shown (Sakurai, 1979; Tsubouchi, 1980) Its broad floodplain and welldeveloped natural levees may have made it more suitable for agriculture and settlement Conclusions We discussed the age and history of the natural levees based on radiocarbon ages in and around the levees and we considered their relationships to delta evolution, and the distribution of archaeological sites During the early Holocene, sediment discharged by the Song Hong enhanced both aggradation of floodplain and river-mouth progradation within the Song Hong drowned valley Radiocarbon dates obtained from cores, trench exposures, and archaeological sites record a very high accumulation rate (aggradation) before cal kyr BP, which coincides with a period of sea-level rise, and a low accumulation rate after cal kyr BP, related to stable and falling sea level Sedimentary records and topography show that the Day River was more active in the middle Holocene during a highstand of sea level, and the present relatively high topography may be relict topography Human settlement along the natural levees of the Song Hong delta started after cal kyr BP, and continued as sea level fell to the present level and the delta prograded rapidly seaward During this period, the longitudinal profiles of the Song Hong and Day rivers were extended and vertical accretion of alluvial sediments ceased in the West Floodplain Settlement sites spread on the lowland as the natural levees formed and extended seaward approximately 2000 years behind the shoreline migration Acknowledgement We thank colleagues in the Vietnam National University, Hanoi, and University of Tokyo for their outstanding help during sampling and analyzing For shell identification, we thank Dr Rei Nakashima in the Geological Survey of Japan, AIST Special thanks to Dr Takashi Kudo for correction of the radiocarbon ages We also thank the editors and anonymous reviewers for constructive comments on our manuscript Funding Financial support for this study was provided by Japan Society for Promotion of Science (JSPS) Asia-Africa Science Platform Mega-delta Watching project References Brierley GJ, Ferguson RJ and Woolfe KJ (1997) What is a fluvial levee? Sedimentary Geology 114: 1–9 Bureau of Flood Control Economic Commission for Asia and the Far East (1951) Methods and Problems of Flood Control in Asia and the Far East United Nations Coleman JM (1969) Brahmaputra River: Channel processes and sedimentation Sedimentary Geology 3: 129–239 Collinson JD (1996) Alluvial sediments In: Reading HG (ed.) Sedimentary Environments: Processes, Facies and Stratigraphy Oxford: Blackwell Scientific Publications, pp 37–82 Funabiki A, Haruyama S, Nguyen VQ et al (2007) Holocene delta plain development in the Song Hong (Red River) delta, Vietnam Journal of Asian Earth Sciences 30: 518–529 Goodbread SL Jr and Kuehl SA (2000) The significance of large sediment supply, active tectonism, and estuary on margin sequence development: Downloaded from hol.sagepub.com at The University of Iowa Libraries on March 19, 2015 648 The Holocene 22(6) Late Quaternary stratigraphy and evolution of the Ganges-Brahmaputra delta Sedimentary Geology 133: 227–248 Gourou P (1936) Le paysans du delta tonkins Les Editions d’art et d’histoire Les Editions d’art et d’histoire, 541 pp Haruyama S, Funabiki A and Le VT (2003) Landform and flood in the northern part of Red River plain Water Science 5: 47: 50–55 (in Japanese) Hori K, Tanabe S, Saito Y et al (2004) Delta initiation and Holocene sea level change: Example from the Song Hong (Red River) delta,Vietnam Sedimentary Geology 164: 237–249 Iseya F and Ikeda H (1989) Sedimentation in coarse-grained sand-bedded meandera: Distinctive deposition of suspended sediment In: Taira A and Masuda F (eds) Sedimentary Facies in the Active Plate Margin Tokyo: Terra, pp 81–112 Jusseret S, Tam VT and Dassargues A (2009) Groundwater flow modelling in the central zone of Hanoi, Vietnam Hydrogeology Journal 17: 915–934 Lam DD (2003) Sedimentary evolution of the Red River Delta during the Holocene PhD Thesis, Vietnam National University, Hanoi, University of Science, Vietnam Lam DD and Boyd WE (2000) Holocene coastal stratigraphy andmodel for the sedimentary development of the Hai Phong area in the Red River delta, north Viet Nam Journal of Geology (Series B) 15–16: 18–28 (in Vietnamese with English abstract) Lam DD and Boyd WE (2003) Holocene coastal stratigraphy and the sedimentary development of the Hai Phong area of the Bac Bo Plain (Red River Delta), Vietnam Australian Geographer 34: 177–194 Lam DD and Boyd WE (2008) Delta lobes of the Red River Delta Journal of Geology (Series A), 308: 9–10 (in Vietnamese with English abstract) Li Z, Saito Y, Matsumoto E et al (2006) Climate change and human impact on the Song Hong (Red River) delta, Vietnam, during the Holocene Quaternary International 144: 4–28 Mackin JH (1948) Concept for the graided river Geological Society of America, Bulletin 59: 463–512 Mathers S and Zalasiewicz J (1999) Holocene sedimentary architecture of the Red River Delta, Vietnam Journal of Coastal Research 15: 314–325 Mathers SJ, Davis J, Macdonald A et al (1996) The Red River Delta of Vietnam British Geological Survey Technical Report WC/96/02 Miall AD (1992) Alluvial deposits In: Walker RG and James NP (eds) Facies Models: Response to Sea level Change Waterloo Ont.: Geological Association of Canada, pp 119–139 Miall, AD (1996) The Geology of Fluvial Deposits Springer Milliman JD, Rutkowski C and Meybeck M (1995) River Discharge to the Sea: A Global River Index LOICZ Core Project Office Nielsen LH, Mathiesen A, Bidstrup T et al (1999) Modelling of hydrocarbon generation in the Cenozoic Song Hong Basin, Vietnam: A highly prospective basin Journal of Asian Earth Sciences 17: 269–294 Nishimura M and Nishino N (2003) Data Collection of the Archeological Sites in the Red River Plain Report on Grant-in-Aid from the Scientific Research Project of the Ministry of Education, Culture, Sports, Science and Technology of Japan (in Japanese with English abstract) Oketani S, Haruyama S and Sieng A (2007) Floodplain characteristics of the Mekong Delta in Cambodia Geographical Review of Japan 80: 195–205 Petter AL and Muto T (2008) Sustained alluvial aggradation and autogenic detachment of the alluvial river from the shoreline in response to steady fall of relative sea level Journal of Sedimentary Research 78: 98–111 Polach HA (1976) Radiocarbon dating as a research tool in archaeology-hopes and limitations In: Barnard N (ed.) Symposium on Scientific Methods of Research in the Study of Ancient Chinese Bronzes and South East Asian Metal and Other Archaeological Artifacts, Proceedings Australian National University, Department of Far Eastern History Pruszak Z, Szmytkiewicz M, Nguyen MH et al (2002) Coastal processes in the Red River delta area, Vietnam Coastal Engineering Journal 44: 97–126 Rangin C, Klein M, Roques D et al (1995) The Red River fault system in the Tonkin Gulf, Vietnam Tectonophysics 243: 209–222 Sakurai Y (1979) A preliminary essay on the reclamation in the ancient Red River Delta Southeast Asian Studies 17: 3–57 (in Japanese with English abstract) Sakurai Y (1980a) The Red River Delta in the tenth century Southeast Asian Studies 17(4): 597–632 (in Japanese with English abstract) Sakurai Y (1980b) The Red River Delta during Ly Dynasty Southeast Asian Studies 18(2): 271–314 (in Japanese with English abstract) Sakurai Y (1989) The Red River Delta in the Tran Dynasty Southeast Asian Studies, 27(3): 275–300 (in Japanese with English abstract) Stuiver M and Polach HA (1977) Discussion: Reporting of 14C data Radiocarbon 19(3): 355–363 Stuiver M, Reimer PJ and Braziunas TF (1998) High-precision radiocarbon age calibration for terrestrial and marine samples Radiocarbon 40: 1127–1151 Syvitski JPM and Kettner A (2011) Sediment flux and the Anthropocene Philosophical Transactions of the Royal Society A – Mathematical, Physical and Engineering Sciences 369(1938): 957–975, doi: 10.1098/ rsta.2010.0329 Tamura T, Saito Y, Sieng S et al (2009) Initiation of the Mekong River delta at ka: Evidence from the sedimentary succession in the Cambodian lowland Quaternary Science Reviews 28: 327–344 Tanabe S, Hori K, Saito Y et al (2003a) Sedimentary facies and radiocarbon dates of the ND (Nam Dinh)–1 core sediments obtained from the Song Hong (Red River) delta, Vietnam Journal of Asian Earth Science 21: 503–513 Tanabe S, Hori K, Saito Y et al (2003b) Song Hong (Red River) delta evolution related to millennium-scale Holocene sea level changes Quaternary Science Reviews 22: 2345–2361 Tanabe S, Saito Y, Vu QL et al (2006) Holocene evolution of the Song Hong (Red River) delta system, northern Vietnam Sedimentary Geology 187: 26–61 Tran DT and Dinh VH (2000) Coastal development of the modern Red River delta Bulletin of the Geological Survey of Japan 51: 276 Tsubouchi Y (1980) Significance of the lowlands in Southeast Asia Southeast Asian Studies 17(4): 593–596 (in Japanese) Umitsu M (1987) Late Quaternary sedimentary environment and landform evolution in the Bengal lowland Geographical Review of Japan 60: 164–178 Umitsu M (1993) Late Quaternary sedimentary environments and landforms in the Ganges delta Sedimentary Geology 83: 177–186 Wu B, Wang G, Ma J et al (2005) Case study: River training and its effects on fluvial processes in the lower Yellow River, China Journal of Hydraulic Engineering, 131: 85–96 Appendix 1.  Calibrated radiocarbon data derived in this study from uncalibrated data of Nishimura and Nishino (2003) Site ID Material Measured age (yr BP) δ13C (‰) Conventional age (yr BP) Calibrated age (2σ range) (cal yr BP) Go Trung Lang Cong1 Lang Cong2 Cai Beo Dong Cho Ma Dong Dong Dau1 Dong Dau2 Go Vuon Chuoi Gi Trua Thong Go Mun Viet Khe1 Viet Khe2 Viet Khe3 Chau Can organic material (early Neolithic age cultural layer) organic material (early Neolithic age cultural layer) organic material (early Neolithic age cultural layer) organic material (early Neolithic age cultural layer) organic material (late Neolithic age cultural layer) organic material (late Neolithic age cultural layer) organic material (early Metal age cultural layer) organic material (early Metal age cultural layer) organic material (early Metal age cultural layer) organic material (early Metal age cultural layer) organic material (early Metal age cultural layer) organic material (late Metal age cultural layer) organic material (late Metal age cultural layer) organic material (late Metal age cultural layer) organic material (late Metal age cultural layer) 4790 ± 50 4900 ± 85 4850 ± 70 5645 ± 115 3840 ± 60 4150 ± 60 2830 ± 80 2960 ± 150 3070 ± 100 2700 ± 90 2430 ± 60 2480 ± 100 2420 ± 100 2330 ± 100 2330 ± 60 -25 ± -25 ± -25 ± -25 ± -25 ± -25 ± -25 ± -25 ± -25 ± -25 ± -25 ± -25 ± -25 ± -25 ± -25 ± 4750 ± 60 4860 ± 90 4810 ± 80 5605 ± 120 3800 ± 70 4110 ± 70 2790 ± 90 2920 ± 150 3030 ± 110 2660 ± 100 2390 ± 70 2440 ± 110 2380 ± 110 2290 ± 110 2290 ± 70 5520 (5440-5590) 5600 (5450-5750) 5550 (5440-5660) 6430 (6180-6680) 4240 (4070-4410) 4670 (4510-4830) 2920 (2750-3080 ) 3080 (2750-3400) 3190 (2930-3450 ) 2730 (2460-2300) 2470 (2310-2620) 2530 (2310-2760) 2510 (2300-2730) 2290 (2040-2550) 2300 (2120-2490) Downloaded from hol.sagepub.com at The University of Iowa Libraries on March 19, 2015 ... of the Song Hong and Day rivers It starts at the Pleistocene terrace and crosses the natural levees of the Day River, adjacent backswamp, and natural levees of the Song Hong This line ends at the. .. the Song Hong Delta plain This line crosses the natural levees of the Song Hong and Day River, and backswamp between two rivers It ends at the backswamp on the left side of the Song Hong On the. .. approximate centre of the backswamp region between the Song Hong and Day rivers, the natural levees of the Song Hong, and ends at the adjacent backswamp on the left side of the Song Hong Late Metal