Seasonal variations of sediment discharge from the Yangtze River before and after
impoundment of the Three Gorges Dam
Kehui Xu , John D. Milliman
School of Marine Science, College of William & Mary, Gloucester Point, VA 23062, USA
a b s t r a c ta r t i c l e i n f o
Article history:
Received 23 May 2008
Received in revised form 5 September 2008
Accepted 8 September 2008
Available online 19 September 2008
Keywords:
Yangtze River
Sediment
Three Gorges Dam
Reforestation
East China Sea
Over the past decades, N50,000 dams and reforestation on the Yangtze River (Changjiang) have had little
impact on water discharge but have drastically altered annual and particularly seasonal sediment discharge.
Before impoundment of the Three Gorges Dam (TGD) in June 2003, annual sediment discharge had decreased
by 60%, and the hysteresis of seasonal rating curves in the upper reaches at Yichang station had shifted from
clockwise to counterclockwise. In addition, the river channel in middle-lower reaches had changed from
depositional to erosional in 2002.
During the four years (2003­2006) after TGD impoundment, ~60% of sediment entering the Three Gorges
Reservoir was trapped, primarily during the high-discharge months (June­September). Although periodic
sediment deposition continues downstream of the TGD, during most months substantial erosion has
occurred, supplying ~70 million tons per year (Mt/y) of channel-derived sediment to the lower reaches of the
river. If sand extraction (~40 Mt/y) is taken into consideration, the river channel loses a total of 110 Mt/y.
During the extreme drought year 2006, sediment discharge in the upper reaches drastically decreased to
9 Mt (only 2% of its 1950­1960s level) because of decreased water discharge and TGD trapping. In addition,
Dongting Lake in the middle reaches, for the first time, changed from trapping net sediment from the
mainstem to supplying 14 Mt net sediment to the mainstem. Severe channel erosion and drastic sediment
decline have put considerable pressure on the Yangtze coastal areas and East China Sea.
Published by Elsevier B.V.
1. Introduction
Global river systems have been increasingly altered by dam
construction and water diversions for water and energy needs
(Nilsson et al., 2005). Humans are simultaneously increasing river
sediment transport through soil erosion and decreasing it through
sediment retention in reservoirs (Syvitski et al., 2005). Fluvial
sediment delivered to the oceans has been estimated to be
~20 billion tons per year (Bt/y) (Milliman and Syvitski, 1992), but
the world's registered 45,000 large reservoirs can effectively trap as
much as 4­5 Bt/y (Vorosmarty et al., 2003). More than 80% of water
and sediment discharge from the Indus River, for instance, has been
diverted by large reservoirs and flow diversion (Giosan et al., 2006).
China has been particularly active in dam construction, building
more than half of the world's large dams commissioned since 1950
(Fuggle and Smith, 2000). Over the past 25 years, for example, the
Yellow River sediment discharge has decreased by ~90% because of
dam construction and increased water consumption (Wang et al.,
2006, 2007b). Of concern in this paper is the impact of the world's
largest dam, the Three Gorges Dam (TGD), on the Yangtze River
(Changjiang) (Fig. 1).
The TGD began to impound water and sediment discharge on 1
June 2003; by 2009 when the TGD is in full operation, the water
storage capacity of the Three Gorges Reservoir (TGR) will be 39 km3
,
about 4.5% of the Yangtze's annual discharge. Because the Yangtze not
only sustains 440 million inhabitants in its drainage basin but also
helps nourish rich fishing grounds in the East China Sea, the response
of the Yangtze lower reaches and coastal area to upstream damming
has raised considerable interest (Shen and Xie, 2004). In the past
decade, for example, the Yangtze delta has changed from one of
progradation to regression and its wetlands have been disappearing
(Yang et al., 2006a); the degree to which primary production in the
East China Sea has been affected, however, is a matter of debate (Gong
et al., 2006; Yuan et al., 2007).
Land use change, most of which predates the TGD, also has affected
the Yangtze drainage basin. Although water discharge has changed
little, sediment discharge in the upper (at Yichang station, Fig. 1) and
lower (at Datong) reaches of the river had decreased N60% before the
impoundment of the TGR in 2003 (Fig. 2). These pre-TGD declines were
mainly due to construction of N50,000 tributary dams, reforestations
and decreased precipitation in the high-sediment-yield basin of the
Jialing River (Xu et al., 2007). Nevertheless, impact after the TGD was
Geomorphology 104 (2009) 276­283
 Corresponding author. Current address: Departments of Marine Science, Coastal Carolina
University, P.O. Box 261954, Conway, SC 29528, USA. Tel.: +1 843 349 6494; fax: +1 843 349
4042.
E-mail address: kxu@coastal.edu (K. Xu).
0169-555X/$ ­ see front matter. Published by Elsevier B.V.
doi:10.1016/j.geomorph.2008.09.004
Contents lists available at ScienceDirect
Geomorphology
journal homepage: www.elsevier.com/locate/geomorph
immediate: during the two years after impoundment (2003­2004),
sediment discharge at Yichang and Datong decreased by 164 and
128 million tons (Mt) (Fig. 2), respectively, and intense riverbed
scouring led to active channel erosion (Xu et al., 2006). This declining
trend in sediment transport was further exacerbated by the extreme
drought in 2006 when flood season water level was lowest in the past
50 years (Dai et al., 2008); water discharge decreased by ~30%, and
sediment drastically decreased to 9 Mt (only 2% of its 1950­1960s
level) in the upper reaches at Yichang (Fig. 2).
During the past decade, numerous studies have been documented
water and sediment discharge of the Yangtze River. Long-term
sediment flux from the river to the sea since the 1860s was
reconstructed using water discharge data collected from the middle
reaches (Wang et al., 2008). Statistical analyses of sediment and runoff
changes at four mainstem and three tributary stations during the past
55 years were done by Zhang et al. (2006). Sediment flux sensitivity to
climate change (e.g., temperature and precipitation) in upper reaches
was documented by Zhu et al. (2008). Flood records in a sediment core
collected in the middle reaches and its sediment sources were
investigated by Yi et al. (2006). Dam impacts on sediment discharge,
channel variations, and delta erosion have been further studied
by Yang et al. (2006a, 2007) and Yang et al. (2006b). Temporal and
spatial variations of basinwide and subbasin discharge of the river
were analyzed by incorporating climatic and anthropogenic impacts
(Xu et al., 2007). In addition, sediment budget analysis and sand
extractions were discussed by Chen et al. (2006, 2008), Walling
(2007), and Wang et al. (2007a).
Although many aspects of water and sediment discharge of
the Yangtze have been addressed, few have delineated (i) seasonal
rating-curve changes before and after the TGR impoundment,
(ii) monthly sediment trapping in the TGR, and (iii) seasonal variations
of channel erosion/deposition. These rating-curve changes may
significantly impact the timing and pattern of hydrological cycle and
sediment transport in the Yangtze basin. Reservoir trapping and
channel erosion also play key roles in determining how much
sediment can be delivered to the Yangtze estuary to maintain the
geomorphology of the Yangtze delta. In addition, quantifying seasonal
water and sediment variations is critical for better management of
watershed water/sediment resources and to ensure a sustainable
development for the Yangtze basin.
Our two recent publications (Xu et al., 2006, 2007) mainly
focused on annual variations using data up to 2004 or 2005,
respectively, but here we report on the most recent (1950 to 2006)
seasonal water and suspended-sediment discharge measured at
both mainstem and tributary stations. Because the Yangtze usually is
divided into upper (above Yichang), middle (Yichang to Hankou) and
lower (downstream of Hankou) reaches (Chen et al., 2001), we
emphasize both upper (Yichang, directly downstream of the TGD)
and lower (Datong, the seaward-most station) reaches. In contrast to
other studies, here we (i) show dramatic hysteresis shifting of rating
curves, from clockwise to counterclockwise upstream before and
after the TGD, (ii) analyze annual and particularly monthly sediment
budgets for channels and lakes in the middle-lower reaches, and
(iii) attempt to quantify the forcings that caused the drastic Yichang
sediment decline to 9 Mt in 2006. Because bedload only represents
5% of total sediment discharge (Chen et al., 2001), it is not addressed
in our paper except when considering sand extraction from the river
channel. Most water and suspended-sediment discharge data are
from the Bulletin of Yangtze Sediment (2000, 2001, 2002, 2003,
2004, 2005, 2006, available in http://www.cjh.com.cn/) or Changjiang
Water Resources Commission.
Fig. 1. The Yangtze River drainage basin. Two grey areas in the southeastern basin are Dongting and Poyang Lakes. Note that Yangtze mainstem delivers water and sediment into
Dongting Lake and flows back to the mainstem (shown in arrows). TGD, Three Gorges Dam.
Fig. 2. Temporal variations of annual water and sediment discharge in upper (Yichang
station) and lower (Datong station) reaches of the Yangtze River in 1950­2006. Bold
vertical dashed lines labeled TGD (Three Gorges Dam) indicate the impoundment in
June 2003. Data are from Changjiang Water Resources Commission. Declined sediment
discharge before 2003 was mainly caused by tributaries damming and reforestation.
277K. Xu, J.D. Milliman / Geomorphology 104 (2009) 276­283
2. Mainstem variations at Yichang and Datong
2.1. Temporal variation
Under the control of summer Asian monsoons, water and sediment
discharge from Yangtze upper (Yichang) and lower (Datong) reaches
both show seasonal patterns. In 1950­1990, more than 70% of water
discharge at Yichang and Datong occurred during summer (MayOctober),
with an average peak in July (Fig. 3). Water discharge at two
stations has shown little change during the past 56 years (Fig. 2), with
peak discharge averaging 80 and 130 km3
/month (mo), respectively
(Fig. 3). In June 2003, about 10 km3
of water impoundment in the TGR
had relatively small impacts on water discharge at two stations.
Interestingly, a decrease in water discharge was obvious at Yichang in
August 2003 (Fig. 3); this decline is probably related to decreased
precipitation, as no impoundment took place at this time.
Sediment discharge at both stations, however, has decreased
dramatically since 1950 (Fig. 2). Peak monthly sediment discharge at
Yichang decreased from 160 Mt/mo in 1950­1990 to only 7 Mt/mo in
2006; Datong peak discharge also declined from 110 to 17 Mt/mo over
the same period (Fig. 3). Sediment decrease is more drastic in Yichang
than that in Datong (Figs. 2 and 3), mainly because of the
compensation from channel erosion downstream of the TGD
(discussed later). The timing of peak sediment discharge at both
stations, however, shifted from July before 2000 to August or
September in 2001, 2002, 2004, and 2005 (July is marked by thin
dashed lines in Fig. 3 for comparison).
2.2. Cumulative discharge variation
The plots of cumulative sediment discharge versus water discharge
(1995­2006) at Yichang and Datong both show stepwise curves, which
were fairly steep in summer but gentle in winter (Fig. 4). The curve of
Yichang began to bend slightly in 2001­2002 and became very flat after
theimpoundmentin June 2003. Incontrast, thecurve atDatong changed
more gradually, showing a much smaller impact by the TGD impoundment
(Fig. 4). Because the slope of the curve indicates sediment
Fig. 3. Monthly water and sediment discharge measured in upper (Yichang) and lower (Datong) reaches from 1950 to 2006. Note that 1950­1990 and 1991­2000 are long-term
monthly means, followed by monthly discharge from 2001 to 2006. For ease of comparison, thin vertical dashed lines represent July for each year; bold vertical dashed lines labeled
TGD (Three Gorges Dam) indicate the impoundment in June 2003.
Fig. 4. Cumulative monthly water and sediment discharge at upper (Yichang) and lower
(Datong) reaches of the Yangtze River in 1995­2006. The inset shows gentle and steep
curves in winter and summer, respectively. TGD (Three Gorges Dam) impoundment in
June 2003 is highlighted for the comparisons before and after the impoundment.
278 K. Xu, J.D. Milliman / Geomorphology 104 (2009) 276­283
concentration, sediment concentrations at Yichang in summer were
much higher than in winter, with concentrations decreasing rapidly
after June 2003.
2.3. Rating-curve variation
Yangtze water mainly comes from the SE owing to higher
precipitation, but Yangtze sediment is from the NW because of more
erodible soils and steep mountains (Xu et al., 2007). Each year Asian
monsoons move northward from the low-sediment-yield SE (in April
and May) to the high-yield NW (June­August) (Shi et al.,1985). In upper
reaches at Yichang during 1950­1990, sediment concentration in
the rising stage (June) on the rating curve was greater than that in the
falling stage (October) presumably because there was more easilyerodible
sediment available in June (Fig. 5). In contrast, lower-reach
variations at Datong reflect the combination of spatial source difference
and temporal monsoon moving. In 1950­1990 Datong's sediment
concentration in the rising stage of water discharge (June, sedimentpoor)
was lower than that in the falling stage (August­September,
sediment-rich) during which turbid sediment was supplied by upper
reaches (Fig. 5). These temporal and spatial settings led to a unique
downstream transition on the hysteresis of rating curves from clockwise
at Yichang to counterclockwise at Datong in 1950­1990 (Fig. 5).
This downstream transition, however, has changed gradually
since the 1990s. The curves at Yichang switched to being counterclockwise
in 1992 and 1995­1997 but remained clockwise in the
other years of the 1990s, leading its 1991­2000 curve to shift to a
semi-counterclockwise hysteresis (Fig. 5). It then fell and became
counterclockwise in 2001­2002, after which it maintained counterclockwise
and dropped further in 2003­2005 and 2006. Of these
decreases in sediment concentrations at Yichang, the most dramatic
is July of 2001­2002 when concentration decreased by 60%
compared with 1991­2000; in contrast, sediment concentration in
August declined by only 20% over the same periods (Fig. 5).
Downstream at Datong, all the curves since 1991 fell below the
1950­1990 line but remained counterclockwise (Fig. 5).
3. Spatial and temporal variations
Variations at Yichang and Datong actually reflect changes on both
mainstem and tributaries in the Yangtze basin. As such, our discussion
of changes on water and sediment discharge will focus on the
transition from the upper to the lower reaches: upstream, inside and
downstream of the TGR, respectively.
3.1. Upstream of the TGR
In 2006, the upper reaches of the Yangtze experienced an extreme
drought, during which water level in flood season was the lowest in
the last 50 years, a situation called "no flood in the flood season" (Dai
et al., 2008). Annual water discharge at Qingxichang (upstream of the
TGR, Fig. 1) also reached its lowest level (278 km3
/y) in the past half
century, ~30% less than its long-term (1956­2005) mean (390 km3
/y)
(Fig. 6). Since the year 1986 was the beginning of the Yangtze
Upstream Water and Soil Conservation Project and 2003 is the first
TGR impoundment, the plots of water versus sediment discharge at
Qingxichang can be divided into three periods: 1956­1985, 1986­
2002, and 2003­2006. Sediment discharge began declining in 1986,
and reached its lowest level in 2006 (Fig. 6). Based on the rating curve
for 2003­2006, decreased water discharge in 2006 led to a drop of
110 Mt in sediment discharge (dashed lines in Fig. 6).
3.2. Sediment trapping inside the TGR
The 600-km-long TGR is located in the upper stream of the
Yangtze, and sediment trapping inside the reservoir has been
substantial since the impoundment in 2003. Located immediately
Fig. 5. The rating curves of monthly water discharge versus sediment concentration for
1950­1990, 1991­2000, 2001­2002, 2003­2005, and 2006 at Yichang and Datong,
respectively. Numbers on the curves represent the sequential months (7 for July 477 and
8 for August) of the years, and arrows show the clockwise or counterclockwise
hysteresis loops. The curves at Yichang shifted from clockwise in 1950­1990 to
counterclockwise after 2001.
Fig. 6. The rating curves of annual water discharge versus sediment discharge at
Qingxichang station in 1956­1985,1986­2002, and 2003­2006. 2006 was an extremely
dry year during which decreased water discharge led to a 110-Mt decline in sediment
discharge. Discharge of Qingxichang was calculated by adding discharge from a nearby
upstream station (Zhutuo) to that of two tributaries (at Beibei and Wulong stations)
between Zhutuo and the TGR.
279K. Xu, J.D. Milliman / Geomorphology 104 (2009) 276­283
upstream and downstream of the TGR, Qingxichang and Yichang
stations have been used to monitor the sediment entering and exiting
the reservoir, respectively (Fig. 1). The contribution from nearby small
rivers to the TGR has been regarded as small, although it may reach as
high as ~25 Mt/y (Yang et al., 2007). During 2002 and before TGD
impoundment, sediment discharge passing two stations was similar,
indicating little net erosion or deposition along the channel between
Qingxichang and Yichang (Fig. 7). After impoundment, however, 124,
102,151, and 93 Mt of sediment were trapped in 2003, 2004, 2005, and
2006, respectively, an average of 118 Mt/y (Fig. 7).
During 2003­2006, sediment retention occurred mainly from June
to September, a pattern similar to the sediment discharge at
Qingxichang (Fig. 7). The more supply from upstream of the TGR,
the more trapping of sediment inside the reservoir. Monthly sediment
discharge entering the TGR correlated with trapped sediment in the
TGR (Fig. 8); approximately 50% of the sediment entering the TGR was
trapped based on the slop of the line in Fig. 8. The highest trapping
(60 Mt) in July 2005 corresponded to the highest sediment discharge
(90 Mt) passing Qingxichang.
Inside the TGR, sediment was mainly accumulated near the
thalweg and became thicker nearer to the TGD; maximum accumulation
thickness (53 m) occurred 6 km upstream of the TGD (Bulletin of
Yangtze Sediment, 2005, 2006). In August 2003, however, sediment
discharge at both Qingxichang and Yichang was unusually lower than
in the preceding and following months (Fig. 7). Because impoundment
in the TGR at that time was not significant, this is likely related to
decreased water discharge (Fig. 3) and reduced precipitation
upstream of the TGR, which propagated downstream to Yichang.
3.3. Channel erosion downstream of the TGR
Another dramatic change in response to TGR impoundment has
been channel erosion downstream of the TGD, which plays a key role
in sediment transfer in the middle and lower reaches. Along the 1140km
stretch between Yichang and Datong, the Yangtze has lost some
sediment (due to siltation and reclamation) through several passages
(small arrows in Fig. 1) connecting Dongting Lake, but has received
considerable sediment from tributaries merging at Poyang Lake and
the Han River. Prior to 2000, sediment transport in this section
displayed extreme seasonal variations, sediment being deposited
during high-discharge months (June­September) but subsequently
eroded in low-discharge months (Fig. 9); net deposition between
Yichang and Datong averaged 48 Mt/y (Table 1). Deposition downstream
of Yichang declined quickly to 15 Mt in 2001, and then
reversed to net erosion (-46 Mt/y) in 2002 (Table 1), erosion occurring
Fig. 7. Monthly sediment discharge at Qingxichang (A) and Yichang (B), and trapped sediment in the TGR (C). The tributary upstream of Qingxichang is Jinsha River. Discharge of
Qingxichang was calculated by adding discharge from a nearby upstream station (Zhutuo) to that of two tributaries (at Beibei and Wulong stations) between Zhutuo and the TGR.
Fig. 8. Suspended sediment discharge entering the Three Gorges Reservoir (TGR) in
2003­2006 correlates with monthly trapped sediment in the TGR. The gauging stations
measuring discharge entering the TGR is Qingxichang.
280 K. Xu, J.D. Milliman / Geomorphology 104 (2009) 276­283
in 10 of 12 months (Fig. 9). In 2002, for the first time, sediment
discharge at Datong exceeded that at Yichang (Table 1). As these
changes occurred before TGD impoundment, clearly human activities
(e.g., damming and reforestation) on the Yangtze tributaries had
collectively played a key role in this deposition-to-erosion transition.
After the 2003 impoundment, channel erosion occurred in every
month in 2003 and 2006 and most months in 2004­2005 (Fig. 4) in
response to the release of water from the TGR with sharply reduced
sediment concentrations. In 2003­2006, an average of ~70 Mt/y of
sediment was eroded between Yichang and Datong, mostly reflecting
the abrupt summertime change from deposition to erosion (Fig. 9).
Channel erosion mainly took place 630 km directly downstream from
the TGD (Yichang to Hankou) but diminished farther downstream
from Hankou to Datong (Figs. 1 and 9; Table 1). Because channeleroded
sediment was generally coarser than suspended sediment,
median grain size increased from 5 m at Yichang to 11 m downstream
at Hankou in 2005 (Bulletin of Yangtze Sediment, 2005).
4. Discussion
During the past half century, human activities have increased
in the Yangtze basin. Since 1986, the Yangtze Upstream Water and Soil
Conservation Project has constructed nearly one million ponds,
reservoirs, and levees, and has increased vegetation cover by 23%;
these land conservation activities alone are believed to be responsible
for the trapping of ~60 Mt of sediment in the upper reaches (Bulletin
of Yangtze Sediment, 2003). By 2000, more than 50,000 dams had
been constructed throughout the Yangtze basin. The Danjiangkou
Dam in the north, for example, has decreased N95% of sediment
discharge from the Han River (Fig. 1) since 1968 (Yang et al., 2006b).
4.1. Shifting of rating curves
These human activities have greatly influenced hydrological
seasonality and hysteresis of the rating curve in the Yangtze River.
Upstream of Yichang, sediment concentration experienced the most
dramatic decrease in July of 2001­2002 (Fig. 5), and peak sediment
discharge shifted from July before 2000 to August­September in
2001­2002 and 2004­2005 (Fig. 3). The July decline played a major
role in the shift of the rating curve from clockwise to counterclockwise.
These drastic sediment drops at Yichang are explained by
the effect of numerous reservoirs in tributaries of the river as follows.
When precipitation is low in January­May, water levels in most
reservoirs fall to meet domestic and agricultural needs. With the reinitiation
of monsoonal summer rains in the upper reaches (usually in
late June and July), reservoirs begin filling, thereby trapping both
sediment and water. Once filled in August­September, water and
sediment again are discharged from reservoirs. One result of reservoir
trapping has been the significant temporal decrease in median grain
size of suspended sediment at Yichang, from 22 m in 1950­2000 to
only 5 m in 2005 (Bulletin of Yangtze Sediment, 2005).
4.2. Sediment budget and complications
Sediment budget provides a way to understand the distribution of
sediment in different parts of a river basin (Wang et al., 2007a).
Fig. 9. Monthly channel deposition (+)/erosion(-) between Yichang and Datong stations, calculated by subtracting Datong from the sum of Yichang, Dongting Lake, Poyang Lake, and
Han River. The net depositional pattern in 1950­2000 gradually shifted to erosional in 2002, and more significantly in 2003­2006. Note that monthly data of Dongting, Poyang, and
Han before 2002 are not available; the pre-2002 plot represents the difference between the discharge at Yichang and that at Datong. Annual variations are shown in Table 1.
Table 1
Sediment budget in the middle-lower reaches of the Yangtze River (unit: Mt/y)
Year Yichang Dongting
Lake
Han
River
Hankou Poyang
Lake
Datong Channel deposition(+)/erosion(-)
Yichang to Hankou (630 km) Hankou to Datong (510 km) Yichang to Datong (1140 km)
1950­2000 501 -86 56 404 10 433 67 -19 48
2001 299 -23 3 285 12 276 -6 21 15
2002 228 -16 3 239 14 275 -24 -22 -46
2003 98 -3 14 165 18 206 -56 -23 -79
2004 64 0 5 136 14 147 -67 3 -64
2005 110 -8 17 174 16 216 -55 -26 -81
2006 9 14 3 58 14 85 -32 -13 -45
Yichang, Hankou and Datong are three mainstem stations from upper to lower streams, and Dongting Lake, Han River and Poyang Lake (all italicized) deliver tributary discharges into
the mainstem (Fig. 1). The Yangtze generally loses sediment into Dongting Lake and receives sediment from Poyang Lake and Han River. Channel deposition(+)/erosion(-) from
Yichang to Datong was calculated by subtracting Datong from the sum of Yichang, Dongting, Poyang and Han. The budgets for Yichang­Hankou (630 km channel length) and
Hankou­Datong (510 km) were calculated in a similar way. All channel budgets are shown in bold. Dongting sediment itself was calculated by subtracting four passages flowing into
Dongting Lake from one passage going back to the mainstem (Fig. 1; Xu et al., 2007). See Fig. 9 for monthly variations.
281K. Xu, J.D. Milliman / Geomorphology 104 (2009) 276­283
Sediment discharge delivered to the lower reaches at Datong is
dependent on (i) sediment discharge at Yichang, (ii) net sediment flux
between Dongting Lake and mainstem, (iii) discharge from un-gauged
small rivers, (iv) contributions from the Han River and Poyang Lake, as
well as (v) sand extraction along channels in the middle-lower reaches
(Fig. 1; Chen et al., 2008). Based on our sediment-budget analysis,
about 70 Mt/y of channel-derived sediment was eroded and delivered
to the lower reaches in 2003­2006 (Table 1). This budget analysis,
however, does not take into account small ungauged rivers in
Yichang­Datong and sand extraction. Considering the low-sediment
yield in the middle-lower reaches, however, contribution from small
ungauged rivers should be fairly small.
Sand extraction in the middle and lower Yangtze basin began as
early as the 1950s. In the 1980s, sand extraction rapidly increased to
40 Mt/y and reached ~80 Mt/y in the late 1990s (Wang et al., 2007a).
In 2003 the Yangtze River Conservation Commission issued a
regulation that total sand extraction/mining should be b34 Mt/y, but
the ban on illegal sand extractions has become a challenge to Chinese
government (Chen et al., 2006). Actual sand extraction in 2003­2006
may have between 34 and ~80 Mt/y. One complicated issue for the
budget analysis here is that extracted sand (N64 m) is much coarser
than the suspended sediment (5­11 m in median size). When sands
are extracted from river bed or banks, there may be more space for
suspended fine sediment to deposit on the bed. If we assume 40 Mt/y
of sand extraction combined with 70 Mt/y channel erosion (Table 1),
the river channel could have lost as much as 110 Mt/y in 2003­2006.
The resulting channel erosion could result in river bank landslides and
levee failures, endangering the flood control and navigation in the
middle-lower reaches.
Dongting Lake is the major lake located in the middle Yangtze
reaches, receiving discharge from the mainstem and southern tributaries,
which then flows back into the mainstem (arrows in Fig. 1). Net
escape of sediment from mainstem to lake has decreased from 86 Mt/y
in 1950­2000 to 16 Mt in 2002 (Table 1). Because of active channel
erosion downstream of the TGR after June 2003, bed level on the
mainstem channel was lowered and sediment escape from the
mainstem to lake essentially stopped in 2004 (0 Mt). During the
extreme drought year 2006, sharply dropped water discharge and TGR
trapping decreased by 110 and 93 Mt sediment (Figs. 6 and 7) in upper
stream, respectively, leading only 9 Mt sediment passing Yichang.
Because of sediment starvation in the mainstem, sediment transport
changed from escaping from the mainstem before 2005 to supplying
net sediment (14 Mt) back to the mainstem in 2006 (Table 1).
The area of Dongting Lake has decreased from 4350 km2
in 1949 to
2623 km2
in 1995 because of extensive reclamation and severe
siltation (Bulletin of Yangtze Sediment, 2000). Dongting Lake has been
the major freshwater source for millions of people in the middle
reaches of the Yangtze. How to maximize the longevity of Dongting
Lake has long been a serious question. This new lake-to-mainstem
transport pattern, however, might alleviate the shrinking of the lake
and facilitate the management of freshwater resources in the middle
reaches.
5. Conclusions
Before the impoundment of the Three Gorges Reservoir (TGR) in
2003, the seasonality of hydrological processes in the Yangtze upper
reaches had been altered by tributary dams and reforestation such
that the hysteresis of the rating curve in the upper reaches shifted
from clockwise to counterclockwise. After the impoundment, 60% of
the sediment entering the TGR was trapped in flood seasons.
Downstream of the TGR, substantial channel erosion happened most
of the time in 2003­2006. In the extreme drought year 2006,
sediment discharge in the upper reaches drastically decreased to
only 2% of 1950­1960s level owing to both reduced water discharge
and TGR trapping. As downstream channel erosion (70 Mt/y) has not
yet counteracted TGR trapping (118 Mt/y), sediment delivered to the
Yangtze estuary will probably continue to decrease, putting pressure
on the Yangtze lower reaches and adjacent coastal areas.
Acknowledgements
This research was supported by the U.S. National Science
Foundation (NSF) and Office of Naval Research (ONR). We thank
Prof. Z.S. Yang (Ocean University of China) for providing part of the
hydrological data used in this study. We are grateful to two
anonymous reviewers for their comments, suggestions, and critical
reviews.
References
Bulletin of Yangtze River Sediment, 2000. Press of Ministry of Water Resources of the
People's Republic of China. website: http://www.cjh.com.cn/.
Bulletin of Yangtze River Sediment, 2001. Press of Ministry of Water Resources of the
People's Republic of China. website: http://www.cjh.com.cn/.
Bulletin of Yangtze River Sediment, 2002. Press of Ministry of Water Resources of the
People's Republic of China. website: http://www.cjh.com.cn/.
Bulletin of Yangtze River Sediment, 2003. Press of Ministry of Water Resources of the
People's Republic of China. website: http://www.cjh.com.cn/.
Bulletin of Yangtze River Sediment, 2004. Press of Ministry of Water Resources of the
People's Republic of China. website: http://www.cjh.com.cn/.
Bulletin of Yangtze River Sediment, 2005. Press of Ministry of Water Resources of the
People's Republic of China. website: http://www.cjh.com.cn/.
Bulletin of Yangtze River Sediment, 2006. Press of Ministry of Water Resources of the
People's Republic of China. website: http://www.cjh.com.cn/.
Chen, Z., Li, J., Shen, H., Wang, Z., 2001. Yangtze River of China: historical analysis of
discharge variability and sediment flux. Geomorphology 41 (2­3), 77­91.
Chen, X., Zhou, Q., Zhang, E., 2006. In-channel sand extraction from the mid-lower
Yangtze channels and its management: problems and challenges. Journal of
Environmental Planning and Management 49 (2), 309­320.
Chen, X., Yan, Y., Fu, R., Dou, X., Zhang, E., 2008. Sediment transport from the Yangtze
River, China, into the sea over the Post-Three Gorge Dam Period: a discussion.
Quaternary International 186 (1), 55­64.
Dai, Z., Du, J., Li, J., Li, W., Chen, J., 2008. Runoff characteristics of the Changjiang River
during 2006: effect of extreme drought and the impounding of the Three Gorges
Dam. Geophysical Research Letters 35, L07406. doi:10.1029/2008GL033456.
Fuggle, R., Smith, W., 2000. Large Dams in Water and Energy Resource Development in
The People's Republic of China (PRC). Country review paper prepared as an input to
the World Commission on Dams, Cape Town. http://www.dams.org.
Giosan, L., Constantinescu, S., Clift, P.D., Tabrez, A.R., Danish, M., Inam, A., 2006. Recent
morphodynamics of the Indus delta shore and shelf. Continental Shelf Research 26
(14), 1668­1684.
Gong, G., Chang, J., Chiang, K., Hsiung, T., Hung, C., Duan, S., Codispoti, L.A., 2006.
Reduction of primary production and changing of nutrient ratio in the East China
Sea: effect of the Three Gorges Dam? Geophysical Research Letters 33, L07610.
doi:10.1029/2006GL025800.
Milliman, J.D., Syvitski, J.P.M., 1992. Geomorphic/tectonic control of sediment discharge
to the ocean: the importance of small mountainous rivers. Journal of Geology 100
(5), 525­544.
Nilsson, C., Reidy, C.A., Dynesius, M., Revenga, C., 2005. Fragmentation and flow
regulation of the world's large river systems. Science 308 (5720), 405­408.
Shen, G., Xie, Z., 2004. Three Gorges Project: chance and challenge. Science 304 (5671),
681b.
Shi, Y.-L., Yang, W., Ren, M.-e., 1985. Hydrological characteristics of the Changjiang and
its relation to sediment transport to the sea. Continental Shelf Research 4 (1­2),
5­15.
Syvitski, J.P.M., Vorosmarty, C.J., Kettner, A.J., Green, P., 2005. Impact of humans on the
flux of terrestrial sediment to the global coastal ocean. Science 308 (5720), 376­380.
Vorosmarty, C.J., Meybeck, M., Fekete, B., Sharma, K., Green, P., Syvitski, J.P.M., 2003.
Anthropogenic sediment retention: major global impact from registered river
impoundments. Global and Planetary Change 39 (1­2), 169­190.
Walling, D.E., 2007. Global change and the sediment loads of the world's rivers.
Proceeding of 10th international Symposium on River Sedimentation, Moscow,
Russia, August 2007, vol. 1. Moscow State University, pp. 112­130.
Wang, H., Yang, Z., Saito, Y., Liu, J.P., Sun, X., 2006. Interannual and seasonal variation of
the Huanghe (Yellow River) water discharge over the past 50 years: connections to
impacts from ENSO events and dams. Global and Planetary Change 50 (3­4),
212­225.
Wang, Z.Y., Li, Y., He, Y., 2007a. Sediment budget of the Yangtze River. Water Resource
Research 43, W04401. doi:10.1029/2006WR005012.
Wang, H., Yang, Z., Saito, Y., Liu, J.P., Sun, X., Wang, Y., 2007b. Stepwise decreases of the
Huanghe (Yellow River) sediment load (1950­2005): impacts of climate change and
human activities. Global and Planetary Change 57 (3­4), 331­354.
Wang, H., Yang, Z., Wang, Y., Saito, Y., Liu, J.P., 2008. Reconstruction of sediment flux
from the Changjiang (Yangtze River) to the sea since the 1860s. Journal of
Hydrology 349 (3­4), 318­332.
Xu, K.H., Milliman, J.D., Yang, Z.S., Wang, H.J., 2006. Yangtze sediment decline partly
from Three Gorges Dam. EOS Transactions 87 (19) 185,190.
282 K. Xu, J.D. Milliman / Geomorphology 104 (2009) 276­283
Xu, K.H., Milliman, J.D., Yang, Z.S., Xu, H., 2007. Climatic and anthropogenic impacts on
the water and sediment discharge from the Yangtze River (Changjiang),1950­2005.
In: Gupta, A. (Ed.), Large Rivers: Geomorphology and Management. John Wiley &
Sons, West Sussex, England, pp. 609­626.
Yang, S.L., Li, M., Dai, S.B., Liu, Z., Zhang, J., Ding, P.X., 2006a. Drastic decrease in sediment
supply from the Yangtze River and its challenge to coastal wetland management.
Geophysical Research Letters 33, L06408. doi:10.1029/2005GL025507.
Yang, Z.S., Wang, H.-j., Saito, Y., Milliman, J.D., Xu, K.H., Shi, G.-y., 2006b. Dam impacts on
the Changjiang (Yangtze River) sediment discharge to the sea: the past 55 years and
after the Three Gorges Dam. Water Resource Research 42, W04407. doi:10.1029/
2005WR003970.
Yang, S.L., Zhang, J., Xu, X.J., 2007. Influence of the Three Gorges Dam on downstream
delivery of sediment and its environmental implications, Yangtze River. Geophysical
Research Letters 34, L10401. doi:10.1029/2007GL029472.
Yi, C., Liu, H., Rose, N.L., Yang, H., Ni, L., Xie, P., 2006. Sediment sources and the flood
record from Wanghu lake, in the middle reaches of the Yangtze River. Journal of
Hydrology 329 (3­4), 568­576.
Yuan, J., Hayden, L., Dagg, M., 2007. Comment on "Reduction of primary production and
changing of nutrient ratio in the East China Sea: effect of the Three Gorges Dam?"
by Gwo-Ching Gong et al. Geophysical Research Letters 34, L14609. doi:10.1029/
2006GL029036.
Zhang, Q., Xu, C.-y., Becker, S., Jiang, T., 2006. Sediment and runoff changes in the
Yangtze River basin during past 50 years. Journal of Hydrology 331 (3­4), 511­523.
Zhu, Y.-M., Lu, X.X., Zhou, Y., 2008. Sediment flux sensitivity to climate change: a case
study in the Longchuanjiang catchment of the upper Yangtze River, China. Global
and Planetary Change 60 (3­4), 429­442.
283K. Xu, J.D. Milliman / Geomorphology 104 (2009) 276­283