RESEARCH ARTICLE Reservoir Mud Releasing May Suboptimize Fluvial Sand

10.1029/2023WR036701
Supply to Coastal Sediment Budget: Modeling the Impact of
Key Points: Shihmen Reservoir Case on Tamsui River Estuary
• Increasingly adopted mud release
strategy is effective to mitigate Yu‐Ta Hsueh1, Fu‐Chun Wu1 , Qinghua Ye2, Steven Y. J. Lai3, and Yinphan Tsang4
reservoir siltation yet suboptimal to
alleviate coastal sediment deficit 1
Department of Bioenvironmental Systems Engineering, National Taiwan University, Taipei, Taiwan, 2Deltares, Delft, The
• Flood‐driven tributary‐sourced sands
Netherlands, 3Department of Hydraulic and Ocean Engineering, National Cheng Kung University, Tainan, Taiwan,
dominate supply to coastal sediment 4
budget yet sand delivery is reduced by Natural Resources and Environmental Management, University of Hawaiʻi at Mānoa, Honolulu, HI, USA
mantling of released muds
• Sand delivery deficit (relative to sand
delivery of clear‐water flood release Abstract Regular release of sediment from reservoir has been increasingly adopted as a strategy for
scenario) increases linearly with degree sustainable management. Here, we use a process‐based morphodynamic model to simulate the estuarine
of bed mud saturation
sediment dynamics impacted by turbidity current venting implemented by the Shihmen Reservoir during three
typhoon events in 2008. Upon validation with the post‐event bathymetries, the model hindcasts reveal that mud
Supporting Information: releasing can be effective in mitigating reservoir siltation, yet may be a suboptimal strategy for alleviating
Supporting Information may be found in
coastal sediment deficit. A vast majority of the released muds were delivered through the estuary and exported
the online version of this article.
to offshore by flood advection, wave dispersion, and tidal flushing. The flood‐driven sands, sourced mainly
from downstream tributaries, were instead the major contributor to coastal sediment budget. However, mud
Correspondence to:
mantling (covering and immobilizing sand deposits by the reservoir‐released muds) reduced sand availability
F.‐C. Wu,
[email protected] and thus sand delivery to the coast. For the present case, 25% of the released muds were deposited along the way,
presence of these mud covers reduced sand delivery by 15%, compared to a hypothetical scenario of clear‐water
Citation: flood releases. The relative sand transport deficit is found to increase linearly with the degree of bed mud
Hsueh, Y.‐T., Wu, F.‐C., Ye, Q., Lai, S. Y. saturation, 1–D/R, with D/R the ratio of single‐event mud deposit to release. Given broad relevance to global
J., & Tsang, Y. (2024). Reservoir mud reservoirs encountering the problems of siltation and coastal sediment deficit, our findings highlight that
releasing may suboptimize fluvial sand
sustainable management needs to look beyond just a bulk amount of sediment, but it is critical to consider how
supply to coastal sediment budget:
Modeling the impact of Shihmen Reservoir different sediment fractions are interacting and impacted by human activities.
case on Tamsui River estuary. Water
Resources Research, 60,
e2023WR036701. https://doi.org/10.1029/
2023WR036701 1. Introduction
Received 11 NOV 2023 Dams around the world impound rivers for water supply, power generation, flow regulation, and flood control.
Accepted 6 MAY 2024 Reservoir behind dam traps sediment carried by the flow and interrupts their continuity, thereby altering the
downstream flow and sediment regimes and reducing the storage capacity and useful life of a reservoir (Kondolf
et al., 2014). Since fluvial sediment is an important source of the coastal sediment budget, the decline of fluvial
sediment supply would lead to coastal erosion and wetland loss, rendering the global estuarine deltas such as the
Mississippi, Rhône, Rhine‐Meuse, Ebro, Nile, Volta, Mekong, and Yangtze at the risk of drowning in the face of
sea level rise (Allison & Meselhe, 2010; Blum & Roberts, 2009; Boateng et al., 2012; Cox et al., 2021; Dunn
et al., 2019; Darby et al., 2020; Dethier et al., 2022; Kondolf et al., 2018; Schmitt & Minderhoud, 2023; Schmitt
et al., 2017; Syvitski et al., 2005; Tu et al., 2019; Warrick et al., 2019; Yang et al., 2014). To alleviate reservoir
siltation and coastal sediment deficit, regular releases of sediment from reservoirs have been increasingly adopted
as a strategy for sustainable management (Besset et al., 2017; Espa et al., 2019; Fruchard & Camenen, 2012;
Kemp et al., 2016; Kondolf & Yi, 2022; Lee et al., 2022; Rovira & Ibàñez, 2007; Sumi, 2008; Wang et al., 2017).
Depending on their main objective, sediment release operations may be categorized as routing and flushing
(Morris 2020a). The objective of routing is to pass the inflowing sediment‐laden floods; flushing is aimed to
remove the previously deposited sediments by hydraulic scour or resuspension (see Figure 1). For example,
© 2024. The Author(s).
This is an open access article under the drawdown sluicing and turbidity current venting belong to the former, whereas empty flushing and pressure
terms of the Creative Commons flushing belong to the latter. Turbidity current venting is only possible when the high‐concentration density
Attribution‐NonCommercial‐NoDerivs current has sufficient momentum and turbulence to sustain sediment in suspension and move as an underflow
License, which permits use and
distribution in any medium, provided the along the bottom (Chamoun et al., 2016).
original work is properly cited, the use is
non‐commercial and no modifications or The dominant type of sediment released from a reservoir depends largely on what was deposited closest to the
adaptations are made. dam (Asaeda & Rashid, 2012; Brandt & Swenning, 1999). Sediments transported and deposited in the reservoir

HSUEH ET AL. 1 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

Figure 1. (a) Definition sketch of reservoir sedimentation. (b)–(c) Routing of inflowing sediment during floods by drawdown
sluicing and turbidity current venting. Drawdown of water surface results in an increase of energy slope and sediment
transport capacity, facilitating the passage of sediment‐laden floods. Turbidity current venting is feasible when the density
current has sufficient momentum and turbulence to sustain sediment motion along the bottom. The high‐level outlet may be
opened to concurrently release the clear water. (d)–(f) Removal of previously deposited sediment by empty flushing and
pressure flushing. Drawdown of water surface during partial emptying stage facilitates hydraulic scour of delta. Free flows
established during full emptying stage further enhance removal of the redeposited delta scour. Pressure flushing occurs when
reservoir level is high and low‐level outlet is opened to release sediment, producing a localized scour cone adjacent to the
outlet. As there is no significant delta scour, removal of sediment by pressure flushing is limited to the area immediately
upstream of the outlet.

exhibit a typical longitudinal gradation (Figure 1a) (Morris & Fan, 1998). The coarsest fractions (gravel and
coarse sand) are transported as bedload and deposited along the topset of the delta in the upper reach of the
reservoir. Finer fractions (fine sand and coarse silt) are transported as bedload and suspended load by hyperpycnal
flows (turbidity currents) or homopycnal flows, and settling out as tapering deposits along the foreset and bot-
tomset of the delta (Lai & Wu, 2021). The finest fractions, mud (silt and clay), are carried farther downstream by
turbid underflows and can reach the dam to form a wedge‐like muddy pond deposit (Schleiss et al., 2016; Toniolo
et al., 2007). For an overview, we compiled 56 global cases of reservoir sediment release collected from the
literature, distributed in 21 countries over six continents (Figure 2). Among all, 84% (=47/56) had mud as their
main type of sediment released, of which 81% (=38/47) used empty flushing as the mode of operation (for a list of
data, see Table S1 in Supporting Information S1). This highlights the fact that the global reservoir sediment
releases are dominated by drawing down the water level to scour the muddy deposits distributed near the dam
(Morris, 2020a).

Reservoir sediment releases can be also a solution to the issue of coastal sediment deficit. Field investigations
revealed the importance of fluvial sediment supply in nourishing the estuarine deltas and wetlands (Allison,
Nittrouer, et al., 2017; Gelfenbaum et al., 2015; Kemp et al., 2016; Nittrouer & Viparelli, 2014; Warrick
et al., 2015, 2019; X. Wu et al., 2021). These field studies indicated that, although mud dominates the total
sediment load delivered to the coastal waters, it is sand that actually accounts for the vast majority of coastal
deposits building landforms and mitigating land losses. Sand is deposited mainly in the intertidal and shallow
subtidal zones, while mud is transported farther to offshore, where a part is deposited on the seabed and resus-
pended by waves during high tides. Sands delivered to the coasts originate from scour of bed downstream of the
dam, or the hillslope and channel erosion of the downstream tributaries (Wang et al., 2017; X. Wu et al., 2021).

HSUEH ET AL. 2 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

Figure 2. Global cases of reservoir sediment release. Operation mode of sediment release is represented by different symbols; main type of sediment released is
represented by different colors; design capacity of the reservoir is represented by different symbol sizes. A total of 56 global cases, compiled from the literature, are
distributed in 21 countries over 6 continents. Among all, 84% (=47/56) had mud as the main type of sediment released, of which 81% (=38/47) used empty flushing as
the mode of operation.

Moreover, in the Mekong Delta the observed estuarine riverbed was mantled by mud layers, which was regarded
as a key control on mobilization of sand from the riverbed (Allison, Dallon Weathers, & Meselhe, 2017). The
effects of mud mantling on sand transport and thus on coastal sediment budget are highly relevant to the success of
delta nourishment, yet these issues still remain to be elucidated.
To gain useful insights, we use a process‐based morphodynamic model to simulate the sediment dynamics
downstream of the dam and assess the impact of reservoir mud releases. We seek to answer the following research
questions. (a) What are the impacts of the reservoir‐released muds on estuarine and coastal sediment budgets?
Specifically, what difference would it make were clear‐water floods, instead of turbidity currents, released from
the reservoir? (b) If the effects of mud mantling do exist, what impact do these reservoir‐released muds have on
fluvial sand transport? To answer question (a), we isolate the effect of mud releasing by simulation of a hypo-
thetical scenario where clear‐water floods were released from the reservoir. Contrasting the results of the mud‐
release scenario and the hypothetical clear‐water release scenario allows us to identify the impact of mud releases
on estuarine and coastal sediment budgets. To answer question (b), we track the fate (spatiotemporal distribu-
tions) of the reservoir‐released muds, and quantify their mantling effect on fluvial sand transport. Lastly, we
discuss the implications of our findings for strategies of sustainable sediment management.

2. Materials and Methods

2.1. Study Case: Shihmen Reservoir Mud Releasing
Shihmen Reservoir is located in Dahan River, a tributary of Tamsui River (Figure 3a). The latter is the largest
estuary system in Taiwan, draining 2,730 km2 and hosting the Taipei Metropolitan Area. The river mouth is
bounded north eastly by the Fisherman's Wharf Causeway, south westly by the North Breakwater of the Taipei
Port, and connected northwestward to the Taiwan Strait (Figure 3b). The littoral zone is subject to natural forcing
such as tide, wave, river discharge, and typhoon (Hsieh et al., 2020). The semidiurnal M2 tide is the principal
constituent of astronomical tide. With a mean tidal range of 2.2 m and a spring tidal range up to 3 m, the estuary is
classified as mesotidal (C. N. Chen & Tsai, 2017; W.‐B. Chen et al., 2015), dominated by the southwestward
flood currents and northeastward ebb currents. Waves and longshore currents are driven by strong northeast

HSUEH ET AL. 3 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

Figure 3. (a)–(b) Location maps: Tamsui River basin and estuarine system. (c) Turbidity current venting from the low‐level
mud sluiceway, and clear‐water flood concurrently released from the high‐level spillway of Shihmen Reservoir during
Typhoon Soulik (2013). (d)–(e) Nested, orthogonal curvilinear grids. Estuary models: large‐domain (yellow grid); small‐
domain (red grid). Wave models: 1—far‐field open seas (red grid); 2—surrounding waters (yellow grid); 3—northern waters
(white grid); 4—littoral zone (green grid). Image sources (b, d, and e) Google Earth; (c) Northern Region Water Resources
Office, Taiwan.

monsoons in winters and weaker southwest monsoons in summers (Su et al., 2021). With a moderate wave
climate (mean wave height of 1.4 m at the Taipei Port), the coastal area is subject to a tide‐dominated mixed
energy regime (Hayes, 1979). Tidal effect reaches upstream to three major tributaries, namely, Dahan, Xindian,
and Keelung Rivers (Figure 3b), with their tidal limits located at Chenglin Bridge (31.3 km from river mouth),
Xiulang Bridge (32.6 km from river mouth), and Jiangpei Bridge (39.4 km from river mouth). The annual mean
flows of Dahan, Xindian, and Keelung Rivers are 40.3, 72.3, and 25.4 m3/s; flood peaks during typhoons can
reach 6,000 m3/s in Dahan and Xindian Rivers, and 1,000 m3/s in Keelung River.

Tamsui River is the primary source of sand supply to the coastal sediment budget (Hong et al., 2000). Prior to
1958, the annual sediment delivery was 9.63 Mm3. After Shihmen and Feitsui Dams were built on Dahan and
Xindian Rivers in 1964 and 1987 (Figure 3a), the annual delivery declined to 1.85 Mm3 (Lin et al., 2007).
Shihmen Reservoir was built with seven outlet facilities, including three high‐level outlets (a spillway; two flood
tunnels), a mid‐level outlet (an irrigation canal), and three low‐level outlets (two power‐plant penstocks; a bottom
outlet). Large capacities of the high‐level outlets and small capacities of the low‐level outlets rendered the
reservoir to discharge clear water and store turbid water during floods (Figure 1c), causing severe siltation
problems (WRPI, 2008). Bathymetric records revealed that the cumulative siltation volume reached 106 Mm3 as
of 2018, which exceeded 1/3 of the reservoir design capacity (NRWRO, 2022). The siltation was mainly
composed of muds, as sand supplies from upstream of the Shihmen Reservoir rarely exceeded 5% of the sediment
inflows (NRWRO, 2021).

For desiltation purposes, since 2008 Shihmen Reservoir has regularly implemented turbidity current venting
during floods (Figure 3c), achieved by retrofitting one power‐plant penstock into a dedicated mud sluiceway (Lee
et al., 2022). Suspended sediment concentrations (SSC) of outflows have since been monitored routinely
(WRPI, 2008). The retrofit project upgraded the capacity of the power‐plant penstock from 70 to 300 m3/s

HSUEH ET AL. 4 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

(NRWRO, 2022), increasing the mean sediment sluicing ratio (= ratio of sediment outflow to inflow) from 17% to
32% (S. T. Hsu & Wu, 2019). The mud releasing events that took place during three typhoons in 2008, Fung‐
Wong (7/24–8/2), Sinlaku (9/10–9/19), and Jangmi (9/24–10/3), are adopted here as the baseline scenario of
hindcast simulation.

2.2. Model Description

We used the process‐based morphodynamic model Delft3D (Lesser et al., 2004) to investigate the estuarine
sediment dynamics in response to reservoir mud releasing. Delft3D consists of Flow, Wave, and Morphology
modules that can be coupled to simulate the processes of fluvial‐tidal‐wave hydrodynamics, sediment transport,
and morphological changes (for details see Text S1 in Supporting Information S1). It has been applied and
validated in various environments (estuaries, coastal waters, inland rivers) over a wide range of timescales, such
as long‐term, centennial, decadal, sub‐decadal, annual, sub‐annual, seasonal and event scales (e.g., Gelfenbaum
et al., 2015; Hibma et al., 2003; Hopkins et al., 2018; Lesser et al., 2004; Luan et al., 2017; Luijendijk et al., 2017;
Meselhe et al., 2016; van der Wegen & Roelvink, 2008; van der Wegen et al., 2008, van der Wegen, Jaffe, &
Roelvink, 2011; G. Wu et al., 2023; Zhu & Wiberg, 2022; Zhu et al., 2017).
The Flow Module can be run in 3D or depth‐averaged 2DH mode. In this study 2DH mode was used, because
significant sediment transports only occurred during typhoon periods when flood flows and turbidity currents
were concurrently released from the reservoir. Under such extreme flow conditions, salinity intrusion and
stratification of SSC in Tamsui River estuary are very limited (C. N. Chen & Tsai, 2017; W.‐B. Chen et al., 2015;
W.‐C. Liu et al., 2007), rendering the 2DH approach an effective one for analyses of sediment budgets, as
suggested by previous studies (Akter et al., 2021; Barrera Crespo et al., 2019; Nardin et al., 2020).
The Flow Module computes the flowfield using the unsteady shallow‐water equation discretized on a rectangular
staggered finite difference grid, and solves the system by an alternate direction implicit time‐integration scheme.
The Wave Module can be coupled to account for the enhanced bed shear stress due to wave‐current interactions
(Fredsøe, 1984; Soulsby et al., 1993), which plays a key role in the nearshore sediment resuspension. The
resulting flowfield is used to compute the sediment transport field. Morphology module then uses this transport
field to update the bed elevation and composition, based on the continuity equations of individual sediment
fractions (Lesser et al., 2004):

∂z(i)
b ∂S(i)
bx
∂S(i)
by
(1 ε) = fmor (D(i) E(i) ) (1)
∂t ∂x ∂y

where superscripts (i) imply that the quantities apply to the ith fraction; z(i)
b is bed elevation; ε is bed porosity; t is
time; D(i) and E(i) are deposition and erosion fluxes between water column and bed surface; S(i) (i)
bx ,Sby ) are bedload
transport rates in (x,y) directions; fmor is the morphological acceleration factor, introduced to deal with the gap
between the hydrodynamic and morphodynamic timescales (see Section 2.3). The sediment mixture is comprised
of non‐cohesive sands (>63 μm) and cohesive muds (<63 μm). Sands are transported as bedload and suspended
load, evaluated using the approach of van Rijn (1993) and van Rijn et al. (2001). Muds are transported as sus-
pended load only, computed with the advection‐diffusion equations and finite volume approximation (Del-
tares, 2014a), where D(i) and E(i) are treated as sink and source terms. For muds, D(i) and E(i) are evaluated via the
Partheniades‐Krone formulation (Partheniades, 1965) (for details see Text S1 in Supporting Information S1).

For each time step ∆t, Equation 1 is used to calculate the fractional changes ∆z(i)
b in each grid cell, the summation
of ∆z(i) (i)
b over i yields the total change ∆zb. Bed composition is then updated by incorporating ∆zb in the bed
surface. As finer grains settle slower than coarser ones, finer fractions tend to be more prevalent in the upper part
of the bed, covering coarser fractions and becoming more susceptible to erosion. To account for this, the erosion
flux E(i) and bedload transports S(i) (i)
bx ,Sby ) of fraction i are set proportional to its availability in the bed surface
(Deltares, 2014a). In this way, the effect of mud mantling on sand erosion and transport is taken into account.
Morphology module also includes a multi‐layer stratigraphy model that keeps track of bed composition by
bookkeeping the order in which sediments were deposited or eroded (for details see Text S1 in Supporting In-
formation S1). In view of the sediment and hydrodynamic characteristics investigated in this study (see

HSUEH ET AL. 5 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

Section 2.3), a surface layer of 0.2 m and 10 bookkeeping layers each of 1 m were chosen, based on those
suggested for similar sediment mixtures and morphodynamics (Dissanayake et al., 2012; Geleynse et al., 2010;
Valencia et al., 2023; van der Vegt et al., 2016; van der Wegen, Dastgheib, et al., 2011).

2.3. Model Setting

For simulations involving typhoons, the computational domain must be large enough to resolve the wind and
atmospheric pressure fields, that is, large‐scale meteorological forcing that affects waves and tides. Two sets of
nested orthogonal curvilinear grid were generated. The first set was used for hydrodynamic and morphodynamic
simulations focused on coastal and estuarine areas (Figure 3d). To save computational costs, a small‐domain
estuary model (red grid) was nested in a large‐domain estuary model (yellow grid). The small‐domain model
covered the estuarine and nearshore areas, using finer grid (cell sizes: 50–350 m) to resolve local flow, sediment
transport, and morphological processes. The large‐domain model was extended to offshore (cell sizes: 150–
800 m), providing hydrodynamic boundary conditions required by the small‐domain model. To cover the
propagation of tidal wave to its full extent, both models share the same river subdomains (cell sizes: 10–100 m),
which used the tidal limits as the landward boundaries (for details see Text S2 in Supporting Information S1).
The second set of grid was used in the Wave Module (Figure 3e). To incorporate the typhoon tracks and swells
that developed and propagated in open seas and to save computational costs, four layers of grid with increasing
resolution were nested to supply boundary conditions layer by layer: (a) far‐field open seas (red grid, cell sizes:
25 km); (b) surrounding waters (yellow grid, cell sizes: 5 km); (c) northern waters (white grid, cell sizes: 0.8–
2 km); (d) littoral zone (green grid, cell sizes: 40–800 m). The littoral‐zone wave models supplied offshore wave
conditions required by the estuary models.
To pair with the grid, bathymetries of three different scales were derived. Bathymetries of river channels
(Figure 4a) were interpolated from the cross‐sectional surveys of 2008. Bathymetries of the nearshore and
estuarine areas (Figure S1a in Supporting Information S1) were derived using the multibeam echo‐sounding
surveys of 2008. Bathymetries of the open seas (Figure S1b in Supporting Information S1) were retrieved
from Taiwan Ocean Data Bank.
Boundary conditions were used to drive the Flow and Wave Modules. During typhoon periods, hourly flood
hydrographs (Figure 5b) were specified on the landward tributary boundaries. Over the seaward boundary of the
large‐domain estuary model, time series of water levels were generated by eight primary astronomical tide
constituents (M2, S2, N2, K2, K1, O1, P1, and Q1). The amplitude and phase angle of each constituent were derived
from the TPXO global tide model (Egbert & Erofeeva, 2002). Over the cross‐shore lateral boundaries, Neumann
boundary conditions were imposed for free development of water‐level and velocity profiles (Roelvink &
Walstra, 2004). For the Wave Module, wind fields at 10 m above sea surface, retrieved from the ERA5 reanalysis
data of European Center for Medium‐Range Weather Forecasts, were used. The ERA5 typhoon wind and at-
mospheric pressure fields (Figure S2 in Supporting Information S1) were also input to the large‐domain estuary
model to account for the tide level rises (or storm surges) caused by the meteorological forcing (Hu et al., 2009;
Zhu & Wiberg, 2022).
During non‐typhoon periods, the tributary boundaries were specified steady annual mean flows. The input
reduction and morphological acceleration techniques (Roelvink & Reniers, 2012) were used to speed up the
computation (for details see Text S3 in Supporting Information S1). Swells propagating from the far‐field open
seas are negligible under calm climate, thus only the littoral‐zone wave model was used, with schematized
representative waves specified on the seaward boundary, executed via a parallel computing approach. Repre-
sentative waves are a set of partitioned waves with equal energy fluxes that best reproduce the longshore sediment
transport by full wave climate (Figure S3 in Supporting Information S1) (Benedet et al., 2016). In addition, the
schematized morphological tide (Figure 5c) and the morphological acceleration factor fmor were also used.
Morphological tide is aimed to reproduce the same morphological changes as the full tidal time series, using a
shorter diurnal tidal period, which best retains the spring‐neap cycle‐averaged residual sediment transports
(Figure S4 in Supporting Information S1) (Lesser, 2009). A value of fmor = 10 was adopted herein, based on
sensitivity tests over a range of values. These techniques were not used in Stage 1 as typhoon events demand a
more detailed account of the processes involved (Luijendijk et al., 2017; G. Wu et al., 2023).

HSUEH ET AL. 6 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

Figure 4. (a) Bathymetries of river channels, interpolated from cross‐sectional surveys of 2008 using an elliptical inverse
distance weighted method (Merwade et al., 2006). The high‐resolution, multibeam echo‐sounding bathymetries of the
validation area (covering a mouth zone and an estuary zone, bounded by thick black lines) were used for validation of
morphological changes. (b)–(c) Observed and simulated bed elevation changes after Scenario 1. Observed main channel had
deeper erosion around the mid bend, while simulated main channel had deeper erosion at the throat. Four common
morphological features were identified, which are numbered by 1–4.

Three sand and two mud fractions were considered in this study: (a) coarse, medium, and fine sands with
D50 = 500, 200, and 100 μm, based on data collected from Tamsui River estuary and its nearshore area
(TRMO, 2015); (b) coarse (8–63 μm) and fine (<8 μm) muds with settling velocities ws = 0.26 and 0.011 mm/s,
determined from sensitivity tests that reproduced the observed SSC and depositional patterns. The effect of
flocculation on settling velocity was not considered since salinity intrusion was negligible during typhoons. These
sediment fractions were used in the bed material and SSC inputs. An initial bed composition was generated with
an iterative approach presented by van der Wegen, Dastgheib, et al. (2011), using the observed spatial data
(TRMO, 2015) as a starting point. The observed time series of reservoir outflow SSC during the mud release
events were used as the mud inputs to the Dahan River. The proportions of coarse and fine fractions in the mud
SSC were 40% and 60%, based on the data sampled from the bottom outlet (WRPI, 2008). Over the landward
boundaries, the time series inputs of sand SSC were specified with the flow versus suspended sand load (Q‐Qs)
rating curves (for details see Text S4 in Supporting Information S1) derived from the long‐term records (1991–
2018) at six gauges on the downstream tributaries (Figure 3a), namely, Sanying Bridge, Sanxia, Hengxi, Xiulang
Bridge, Baoqiao and Wudu. The proportions of coarse, medium, and fine sands in the SSC were also determined

HSUEH ET AL. 7 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

Figure 5. (a) Flowchart of scenario simulations. Mud release events during three typhoons were simulated consecutively in
Stage 1, followed by tidal cycles during non‐typhoon periods in Stage 2. (b) Hourly water inflow boundary conditions used in
Stage 1. The 21‐day duration encompassed three typhoons (19 days) and two 1‐day flow transitions in between.
(c) Morphological tide used in Stage 2. The 103‐day duration was schematized by 100 cycles of morphological tide. During
Stage 2, water inflows from tributary boundaries were specified steady annual mean flows.

with the iterative approach (van der Wegen, Dastgheib, et al., 2011). The SSC on the seaward boundaries were set
to be zero since they were negligible compared to the fluvial SSC inputs (Lin et al., 2017). The calibrated
parameter values of the Flow/Morphology Modules are summarized in Table S2 in Supporting Information S1.
For the remaining parameters (including those of the Wave Module), the default values were used (Del-
tares, 2014a, 2014b).

2.4. Scenario Simulations

Three scenario simulations were performed. Scenario 1 is a baseline scenario aimed to mimic the mud release
events that took place during three typhoons in 2008. The hindcast results may as well be used for model vali-
dation. Scenario 2 is a hypothetical scenario where clear‐water floods (of zero SSC) were released from the
reservoir, with the same hydrographs as used in Scenario 1. This is equivalent to 100% trapping of sediment
inflow to the reservoir. Contrasting the results of Scenarios 1 and 2 allows us to unravel the effect of mud releasing
on coastal sediment budget. Scenario 3 is a tracking scenario where the reservoir‐released muds were distin-
guished from the preexisting ones, achieved by including two additional mud fractions. The mud fractions
released from the reservoir were identical to the preexisting ones but were assigned new codes, that is, mud_Res_c
and mud_Res_f for the coarse and fine muds. As a result, a total of seven fractions (three sands, two preexisting
muds, and two additional muds) were used in this scenario. Scenario 3 enables us to track the fate of the reservoir‐
released muds, and quantify their impact on fluvial sand transport.

The flowchart of scenario simulations is shown in Figure 5a, the model setup is summarized in Table 1. Mud
release events during three typhoons were simulated consecutively in Stage 1 (21 days), followed by tidal cycles

HSUEH ET AL. 8 of 27
 Table 1
Model Setup for Scenario Simulations

HSUEH ET AL.
Module Stage Model setup Values or data used
Flow/Morphology Stage 1 (typhoon periods) Grid 2 nested layers (with different scales and sizes)
Typhoon wind field and atmospheric pressure field ERA5 reanalysis data (input to large‐domain estuary model for storm surge effects)
Time step 0.1 min
Flow spin‐up interval 1 day
Morphology spin‐up interval 1 day
Duration of flow simulation 21 days
Duration of morphology simulation 19 days (= duration of three typhoon events)
Tributary inflow discharges Hourly flow data
Tributary inflow sand SSC Rating curves
Mud inflow to Dahan River Scenarios 1 and 3: reservoir outflow mud SSC
Scenario 2: none (clear‐water flood releases)
Tide level on seaward boundaries Time series of tide level generated by eight primary astronomical tide constituents
Water level and velocity on cross‐shore boundaries Neumann boundary conditions
SSC on seaward boundaries 0 g/L
Stage 2 (non‐typhoon periods) Grid 2 nested layers (with different scales and sizes)
Time step 0.2 min
Flow spin‐up interval 1.03 days (1 cycle of morphological tide)
Morphology spin‐up interval 4.12 days (4 cycles of morphological tide)
Water Resources Research

Duration of flow simulation 10.3 days (10 cycles of morphological tide)

Morfac fmor (duration of morphology simulation) 10 (103 days = 10 times the duration of flow simulation)
Tributary inflow discharges Annual mean flows (Dahan: 40.3 m3/s; Xindian: 72.3 m3/s; Keelung: 25.4 m3/s)
Tributary inflow sand SSC Rating curves
Mud inflow to Dahan River None (no mud releases)
Tide level on seaward boundary Repeated cycles of morphological tide
Water level and velocity on cross‐shore boundaries Neumann boundary conditions
SSC on seaward boundaries 0 g/L
Wave Stage 1 (typhoon periods) Grid 4 nested layers (with different scales and sizes)
Wave climate Typhoon wind field (ERA5 reanalysis data)
Flow coupling interval 30 min
Stage 2 (non‐typhoon periods) Grid 1 layer (littoral‐zone wave model)
Wave climate 5 representative waves (run in parallel)
Flow coupling interval 30 min
10.1029/2023WR036701

9 of 27
19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

during non‐typhoon periods in Stage 2 (103 days). This two‐stage simulation strategy was adopted for two
reasons. First, as stated above, four layers of wave models were used in Stage 1 while only one layer was used in
Stage 2. Second, the schematization/acceleration techniques were used in Stage 2 but not in Stage 1. Use of
different model setups and computational techniques in Stages 1 and 2 motivated us to perform these simulations
separately. Such strategy is most useful when substantial differences in the magnitude and rate of morphological
change are present between different stages (Roelvink & Reniers, 2012).
Prior to Stage 1, a 24‐hr flow spin‐up and a 24‐hr morphology spin‐up without morphological updating were
performed. Stage 1 included three typhoons total of 19 days, and two 1‐day transition periods in between. This 1‐
day period was deemed sufficient for smooth transitions of flow and SSC, during which most residual suspended
muds would settle out since the estuarine baroclinic circulation induced by salinity intrusion was not considered in
2DH model, lacking the drive for sustaining muds in suspension. Morphology module was turned off during flow
transitions, so the bed topography and composition remained unchanged. After Stage 1, the output bathymetry
and bed composition were input to Stage 2, whose simulation was performed for 100 cycles of morphological tide
(103 days). The total duration of Stages 1 and 2 was 124 days, consistent with the time interval between the pre‐
and post‐event bathymetric surveys that were conducted in June and October 2008. The time steps used in Stages
1 and 2 were 0.1 and 0.2 min, based on a series of sensitivity tests aimed to secure numerical stability and ac-
curacy, that is, the CFL condition. The Wave and Flow Modules were coupled every 30 min. The computation
time required for a scenario simulation was 72 hr (48 and 24 hr for Stages 1 and 2) using an Acer Altos P30 F6
workstation.

2.5. Model Validation

The hydrodynamics, SSC, and morphological changes during typhoon and non‐typhoon periods were used for
model validation. The water levels, velocities, wave heights, and SSCs used were from 14 gauging stations
(locations see Figures 3a and 4a). These include: (a) 3 tide gauges in the northeast and southwest coasts of the
river mouth; (b) 6 flow gauges, 2 in Tamsui River, 1 in Dahan River, 1 in Xindian River, and 2 in Keelung River;
(c) a velocity/SSC gauge in Tamsui River; (d) 4 wave gauges along the northern nearshore. The high‐resolution,
multibeam echo‐sounding bathymetries of the validation area were used for morphological changes, this area
covers a mouth zone and an estuary zone (Figure 4a). The former is bounded by the 20 m isobath, Fisherman's
Wharf Causeway, and North Breakwater of the Taipei Port; the latter is a 40° bend connected to the mouth via a
700‐m‐wide throat and bounded upstream by a cross‐section near Guandu Bridge. In addition, 3 cross‐sections
from the lower, mid, and upper bends (locations see Figures 4b and 4c) were used for validation of the trans-
verse morphological changes.

Three statistical metrics used to assess the model performance include the bias, root mean square error (RMSE)
and Willmott skill score, defined as follows:

1 n
Bias = ∑i=1 (Xmod,i Xobs,i ) (2)
n
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
1 n
RMSE = ∑ (X Xobs,i )2 (3)
n i=1 mod,i
n
∑ |Xmod,i Xobs,i |2
i=1
Skill = 1 n ⃒ ⃒ ⃒ ⃒2 (4)
∑ ⃒Xmod,i Xobs ⃒ + ⃒Xobs,i Xobs ⃒)
i=1

where Xmod and Xobs = modeled and observed values; Xobs = mean observed values; n = number of observed
values. A positive or negative bias indicates a trend of overestimation or underestimation. The RMSE represents
the mean deviation of the modeled from the observed value. A skill score of 1 indicates perfect model perfor-
mance, skills ranging 1–0.65, 0.65–0.5, and 0.5–0.2 indicate excellent, very good, and good model performances,
while poor performances for <0.2 (Willmott, 1981). For hydrodynamic and morphodynamic simulations in the
estuarine and coastal areas, the skill scores of water level and velocity are usually higher than those of wave
height, SSC, and morphology due to greater uncertainties present in the computations of waves and sediment

HSUEH ET AL. 10 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

Table 2
Statistical Metrics of Model Performances
Statistical metrics
Category Validation data (event or time period) Station (or site) Bias RMSE Skill
Tide Water level (non‐typhoon period, 6/10–7/9, 2008) Linshanbi 0.03 (m) 0.11 (m) 0.99
Tamsui River mouth 0.01 (m) 0.10 (m) 1.00
Zhuwei 0.02 (m) 0.10 (m) 1.00
Flow Water level (Typhoon Sinlaku, 9/10–9/19, 2008) Linshanbi 0.00 (m) 0.13 (m) 0.99
Tamsui River mouth 0.05 (m) 0.13 (m) 0.99
Zhuwei 0.03 (m) 0.14 (m) 0.99
Tudigonbi 0.10 (m) 0.16 (m) 0.99
Taipei Bridge 0.07 (m) 0.17 (m) 0.99
Xinhai Bridge 0.09 (m) 0.23 (m) 0.99
Zhongzheng Bridge 0.10 (m) 0.18 (m) 1.00
Bailing Bridge 0.25 (m) 0.39 (m) 0.96
Dazhi Bridge 0.32 (m) 0.55 (m) 0.93
Cross‐sectional velocity (non‐typhoon period, 2008/7/3, semi‐hourly data over a Guandu Bridge 0.05 (m/s) 0.19 (m/s) 0.97
semidiurnal tidal cycle) Taipei Bridge 0.02 (m/s) 0.11 (m/s) 0.99
Xinhai Bridge 0.00 (m/s) 0.07 (m/s) 0.98
Zhongzheng Bridge 0.08 (m/s) 0.14 (m/s) 0.75
Bailing Bridge 0.03 (m/s) 0.13 (m/s) 0.96
Wave Significant wave height (Typhoon Sinlaku, 9/10–9/20, 2008) Taipei Port 0.21 (m) 0.33 (m) 0.97
Hsinchu Buoy 0.15 (m) 0.40 (m) 0.95
Longdong Buoy 0.01 (m) 0.59 (m) 0.97
Guishandao Buoy 0.40 (m) 0.62 (m) 0.95
Sediment Transport SSC (Typhoon Fung‐Wong, 7/24–8/2, 2008) Guandu Bridge 0.31 (g/L) 0.60 (g/L) 0.71
SSC (Typhoon Sinlaku, 9/10–9/19, 2008) 0.62 (g/L) 1.19 (g/L) 0.83
SSC (Typhoon Jangmi, 9/24–10/3, 2008) 0.68 (g/L) 0.97 (g/L) 0.77
Morphological Change Bed elevation change Entire validation area 0.04 (m) 0.57 (m) 0.65
Cross‐sectional bed profile change Lower‐bend (XS‐LB) 0.15 (m) 0.63 (m) 0.98
Mid‐bend (XS‐MB) 0.25 (m) 0.41 (m) 0.99
Upper‐bend (XS‐UB) 0.30 (m) 0.87 (m) 0.98

transports (Zhu & Wiberg, 2022). As such, a skill score of 0.7 for wave height and SSC, and 0.65 for morphology
are normally perceived as reasonably excellent performances. The statistical metrics of all validation results are
summarized in Table 2. To see the outcomes in terms of water level, velocity, wave height, and SSC, the readers
are referred to Text S5 and Figures S5–S9 in Supporting Information S1 therein.

In view of the skill scores, overall, the model performances are excellent in reproducing the tide and water levels,
flow velocities, wave heights, and SSC observed during typhoon and non‐typhoon periods. The RMSE of tide
levels are <5% of the mean tidal range; the RMSE of river flood levels are O (10 1 m) at most gauges. Both of
these exhibit a slight trend of underestimation. The estuarine velocity reversal and asymmetric flow over a tidal
cycle were well captured, with the RMSE being O (10 2–10 1 m/s). The modeled and observed wave heights
during typhoon exhibit good agreement, with the RMSE at the Taipei Port <25% of the mean wave height. Wave
height and estuarine flow velocity both exhibit a slight trend of overestimation. For SSC, the skill scores ranging
0.71–0.83 indicate a reasonably excellent model performance. The sustained high SSC (Figures S9a–S9c in
Supporting Information S1) observed during the post‐peak low‐flow periods of Sinlaku and Jangmi and during
lower floods of Fung‐Wong are attributed to the well‐known local turbidity maximum present near the Guandu
Bridge (W.‐B. Chen et al., 2015; M.‐H. Hsu et al., 2006; W.‐C. Liu et al., 2022, 2007). This local turbidity
maximum is induced by salinity intrusion and estuarine baroclinic circulation that occur particularly during low

HSUEH ET AL. 11 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

flows, trapping suspended sediment in the Guandu deep channel (see Figure 4a and Figure S9d in Supporting
Information S1 for locations). Although such locally sustained high SSC during low flows were not reproduced
by the 2DH model, its effects on the model domain would be minimal given the tributary inflow flood hydro-
graphs used in Stage 1 and the moderate annual mean flows used in Stage 2 (see Section 2.3).

Figures 4b and 4c show the observed bed elevation change and the simulated result from Scenario 1. Overall, the
simulated result is in satisfactory agreement with the observed, except that the observed main channel had deeper
erosion around the mid bend while the simulated result had deeper erosion at the throat. Four common features are
identified. (a) A fan‐shaped deposit off the mouth toward the seaward boundary, attributable to the swinging of
exiting jet flows over tidal cycles during typhoons (Figure S10 in Supporting Information S1). Moreover, waves
exerted additional bed shear stress causing resuspension and transport of bottom sediments, spreading out an
otherwise concentrated mouth deposit (Figure S11 in Supporting Information S1). (b) On the south of the fan‐
shaped deposit was a subzone of scattered erosion and deposition (bounded by North Breakwater and land
boundary). This was also attributed to the wave actions as the wave‐induced bed shear stress resuspended and
transported the otherwise evenly deposited sediments. (c) At the lower bend, substantial amounts of outer‐bank
erosion and inner‐bank deposition were observed. (d) At the narrow upper bend, distinct scour of the main
channel, as much as 1.5 m, was present.
To facilitate further validation, we present areal and volumetric histograms of the observed and simulated bed
elevation changes (Figure S12 in Supporting Information S1). Overall, the simulated and observed histograms
are in satisfactory agreement. The simulated total volume change (= sum of absolute volume changes) and net
volume change (= deposition volume—erosion volume) were smaller than the observed ones by 10%–20%.
Such trend of underestimation was reflected by a negative bias of 0.04 m; the RMSE of 0.57 m implied
mean relative errors of 16% and 14% for total and net volume changes. We looked into the histograms of
individual zones, and found that these errors arose mainly from the estuary zone, due in part to the un-
certainties in the Q‐Qs rating curves used for tributary sand SSC inputs, and factors not considered in this
study (see Section 4.3). Nevertheless, a cell‐by‐cell comparison of the simulated and observed bed elevation
changes yields a skill score of 0.65, suggesting a reasonably excellent model performance. We also compared
the observed and simulated transverse bed profiles (Figure S13 in Supporting Information S1) along three
cross‐sections in the lower, mid, and upper bends (see Figures 4b and 4c for locations). The skill scores
ranging 0.98–0.99 indicate an excellent model performance. In sum, the model hindcast reproduced satis-
factorily the observed morphological changes.

3. Results
3.1. Scenario 1: Baseline Mud Release Events
Based on the simulation results from typhoon and non‐typhoon periods, we derived the stage‐wise maps of bed
elevation change (Figures 6a and 6b), where the cumulative sediment transports and zonal sediment budgets are
also shown. Below, we present separately the results of Stages 1 and 2.

3.1.1. Stage 1—Typhoon Periods

The morphological change during Stage 1 (Figure 6a) highly resembled the full morphological change of Sce-
nario 1 (Figure 4c), indicating that the morphologic development mainly took place in Stage 1. Moreover, the
cumulative sediment transports (white arrows) were all seaward, suggesting that the morphological change was
dominated by fluvial sediment supplies during typhoons. In total, 6.96 M ton of sediment was delivered to the
estuary zone, of which 82% (5.68 M ton) was further transported to the mouth zone, rendering a net budget of
+0.72 Mm3 in the estuary zone. The total input to the mouth zone, 6.13 M ton, was the sum of fluvial supply
(5.68 M ton) and wave‐induced longshore drift (0.45 M ton) (see Figures S11a and S11c in Supporting Infor-
mation S1 for longshore currents and the induced scour). The amount of sediment exported to offshore, 4 M ton,
was 65% of the total input to the mouth zone; the resulting net budget, +1.35 Mm3, was nearly twice the budget in
the estuary zone.

Breaking down the budgets of individual typhoon events, we present in Figure 6c the event‐wise zonal sedi-
ment budgets. Sediment budgets in the two zones were positive during Stage 1. In either zone, Sinlaku played
the most pivotal role, contributing >70% of the total budgets due to the longest flood duration (Figure 5b);

HSUEH ET AL. 12 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

Figure 6. Simulation results of Scenario 1 (baseline). (a–b) Bed elevation changes, cumulative sediment transports, and zonal
sediment budgets during Stage 1 (typhoon periods) and Stage 2 (non‐typhoon periods). White arrows indicate directions of
sediment transport by fluvial flows and tidal currents; black arrows indicate directions of sediment transport by wave‐
induced longshore currents. (c) Event‐wise sediment budgets in mouth and estuary zones. (d) Sand and mud budgets in
mouth and estuary zones during typhoon and non‐typhoon periods.

Jangmi, despite the largest peak, contributed <25% of the total budgets due to its short duration; Fung‐Wong
accounted for <10% of the total budgets due to the smallest flood magnitude and duration. Echoing the trend
presented in Figure 6a, the event‐wise sediment budgets in the mouth zone were consistently greater than those
in the estuary zone.
Breaking down the contributions of sand and mud fractions, we show in Figure 6d the fractional zonal sediment
budgets. In either zone, sand was the major contributor to the sediment budget. Sand contributed 95% of the
budget in the mouth zone and 81% of the budget in the estuary zone (see Text S6; Figures S14a and S14b in
Supporting Information S1). As no sand was released from the reservoir, these sand budgets were mainly sourced
from hillslope and channel erosion of the tributaries. During Stage 1, 3.62 M ton of sand and 3.34 M ton of mud
were input by river flows, 0.35 M ton of sand and 0.09 M ton of mud were input by longshore currents. The total
inputs of sand and mud, 3.97 and 3.43 M ton, differed a little, yet sands and muds exported to offshore, 0.83 and
3.17 M ton, were of notable difference. Specifically, 80% of sand input was deposited yet 92% of mud input was
exported. Of the total sand deposit (1.87 Mm3), 70% was in the mouth zone, whereas only 30% of the mud deposit
(0.21 Mm3) was in the mouth zone. These results corroborate the earlier observation, that is, the fan‐shaped
deposit off the Tamsui River mouth was mainly composed of sandy river‐borne sediments transported by
floods (Hong et al., 2000). In the estuary zone, the mud deposit (0.14 Mm3), despite less than 1/4 of the sand
deposit (0.59 Mm3), revealed that during the release events muds did make a small contribution to the estuarine
sediment budget (see Section 3.3).

HSUEH ET AL. 13 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

3.1.2. Stage 2—Non‐Typhoon Periods

Due to weak bed shear stress during non‐typhoon periods (Figure 6b), the total sediment supply (0.06 M ton) was
two orders of magnitude smaller than that of Stage 1 (6.96 M ton), among which sand supply 0.002 M ton (Figure
S14c in Supporting Information S1) was three orders of magnitude smaller, mud supply 0.06 M ton (Figure S14d
in Supporting Information S1) was two orders of magnitude smaller. During Stage 2, sediments deposited in the
estuary zone during typhoon periods were resuspended and transported by tidal currents. Due to tidal asymmetry
(W.‐C. Liu et al., 2022), seaward sediment transport during ebb tides was greater than landward transport during
flood tides, leading to a seaward residual transport and a negative budget 0.16 Mm3 in the estuary zone, of which
60% was mud and 40% was sand (Figures 6c and 6d). The residual transport to the mouth zone, 0.29 M ton (with
70% mud), plus the input of longshore drift, 0.2 M ton (with 85% sand), was mostly (80%) exported to offshore.
The net budget in the mouth zone, +0.06 Mm3 (with +0.11 Mm3 sand and 0.05 Mm3 mud), was <1/20 the net
budget of Stage 1 (+1.35 Mm3).

3.1.3. Summary of Key Findings

Model hindcast reveals that the flood‐driven sand, sourced from hillslope and channel erosion of the downstream
tributaries, was the major contributor to the estuarine and coastal sediment budgets. During typhoon periods, a
total of 4.19 M ton of sand was supplied from the downstream tributaries (Figure 3a), of which 73% was from
Xindian and Jingmei Rivers, 26% was from Dahan, Sanxia, and Hengxi Rivers, 1% was from Keelung River; 86%
of these sand supplies reached the estuary zone. In the wave‐affected mouth zone, sand accounted for 98% of the
total sediment deposit, where 92% took place during typhoons. Despite a vast amount of mud released from the
reservoir, muds transported to the mouth zone were almost entirely (>97%) exported to offshore by flood
advection and wave dispersion. These results coincide with the observations from the world's mouth bar deposits,
which remain sandy in spite of the mud‐dominated sediment supply (see Braat et al. (2023), van der Vegt
et al. (2020), and references therein). In the tide‐dominated estuary zone, sand also contributed >90% of the
sediment deposit. During non‐typhoon periods, tidal currents resuspended and transported seaward those sedi-
ments deposited during typhoons. The amount of mud so eroded was 50% more than the amount of sand so
eroded. Observations on the flushing of estuarine mud deposits by tidal currents have been reported by, for
example, Guo et al. (2014) and Braat et al. (2017).
These results motivated us to seek answers to the research questions posed in Section 1. To be specific, we aim to
unravel the mantling effect of the reservoir‐released muds on transport of sands. In the following sections,
Scenario 2 isolates the effect of mud releasing, while Scenario 3 tracks the fate of the reservoir‐released muds.

3.2. Scenario 2: Hypothetical Clear‐Water Flood Releases

Two outcomes of Scenario 2 deserve special attention. First, bed elevation changes of Scenario 2 were very
similar to those of Scenario 1 (Figures S14e–S14f in Supporting Information S1). Cell‐by‐cell comparison yields
a coefficient of determination R2 = 0.99, indicating highly coherent morphological changes. Hence, whether
muds were included in the reservoir‐released floods did not affect substantially the resulting morphologies. While
such outcome confirmed the result noted earlier (i.e., sand was the major landform builder), second, the zonal
sediment budgets of Scenarios 1 and 2 did exhibit distinct differences. Relative to Scenario 1, sediment budgets of
Scenario 2 increased by 0.193 Mm3 in the mouth zone yet decreased by 0.045 Mm3 in the estuary zone. To look
into the details, we show in Figure 7 the zonal sand and mud budgets from Scenarios 1 and 2. An apparent trend
emerged, that is, Scenario 2 had greater sand budgets yet negative mud budgets. Specifically, for Scenario 2, the
increase of total sediment budget in the mouth zone was attributed to the larger surplus of sand budget over the
deficit of mud budget, while the reduction of total budget in the estuary zone was due to the larger deficit of mud
budget over the surplus of sand budget.
The deficits of mud budget with Scenario 2 are no surprise, as no muds were released from the reservoir and only
muds preexisting in the downstream channels were eroded by floods. To explain the surpluses of sand budget with
Scenario 2, we show in Figure 8 the post‐Stage 1 bed mud contents of Scenarios 1 and 2. Throughout the estuary
system, bed mud contents of Scenario 2 were invariably lower. For example, at three selected cross‐sections
(Taipei and Guandu Bridges, and River Mouth), bed mud contents of Scenario 2 were 12%, 8%, and 5% lower
than the corresponding ones of Scenario 1. Lower mud contents meant higher sand contents, which in turn led to
higher sand transport rates (as described in Section 2.2) and thus larger sand budgets. Based on the results of

HSUEH ET AL. 14 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

Scenario 2, the smaller sand budgets of Scenario 1 are thus attributable to the
mantling effect of the reservoir‐released muds on sand transports, as reported
by a number of researchers. For example, field observations from the Mekong
Delta revealed that mobilization of bottom sands and suspended and bedload
sand transports were restricted by estuarine mud mantling (Allison, Dallon
Weathers, & Meselhe, 2017; Stephens et al., 2017). Field studies of the
Scheldt estuary showed that up to 40% of the intertidal area was mantled by
muds, limiting the transport of sands that made up 95% of the estuary sedi-
ment volume (Braat et al., 2023). A numerical study applying the Delft3D
model on an idealized estuary revealed that as the mud supply increased,
larger proportions of the bed were mantled by muds whose contents would
rise to a mud‐dominated regime, as a result sands became even less erodible
(Braat et al., 2017).
To quantify the effect of mud mantling, we present in Figure 9 the scenario‐
wise results during typhoons at three selected cross‐sections. Figures 9a–9c
Figure 7. Zonal sand and mud budgets resulting from Scenarios 1 and 2 are flow velocities (with positive and negative values indicating seaward
(baseline and clear‐water release scenarios). Compared to Scenario 1, and landward), where peak and post‐peak periods of Sinlaku and Jangmi were
Scenario 2 had greater sand budgets yet negative mud budgets. For Scenario identified based on the rise/fall of velocities. Figures 9d–9f are the simulated
2, the increase of total sediment budget in mouth zone was attributed to the mud SSC compared to the SSC observed at the bottom outlet. As shown, the
larger surplus of sand budget over the deficit of mud budget, yet the
reservoir‐released mud SSC remained high even after the flood peaks, which
reduction of total budget in estuary zone was attributed to the larger deficit of
mud budget over the surplus of sand budget. sustained the rise of the simulated mud SSC during the post‐peak periods,
particularly evidential at Taipei and Guandu Bridges. The decreasing velocity
and increasing mud SSC during the post‐peak periods led to deposition of muds and increase of bed mud contents
(Figures 9g–9i). Without mud releases (Scenario 2), the rise of bed mud contents during the post‐peak periods
would not occur (see black lines).

Figure 8. Post‐Stage 1 bed mud contents resulting from (a) Scenario 1 (baseline mud‐release scenario), and (b) Scenario 2
(hypothetical clear‐water release scenario). Throughout the estuary system, bed mud contents of Scenario 2 are consistently
lower. Shown here are the mainstem Tamsui River and lower Keelung River, and three cross‐sections used for analyses
(Taipei Bridge, Guandu Bridge, and River Mouth).

HSUEH ET AL. 15 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

Figure 9. (a–i) Simulation results in typhoon periods at three cross‐sections: velocities, scenario‐wise mud suspended sediment concentrations and bed mud contents. (j–
k) Suspended‐load and bedload sand transport deficits (= differential sand transports between Scenarios 1 and 2). Positive and negative values indicate seaward and
landward.

Sand transports were greater without mud releases, either seaward or landward. Figures 9j and 9k show the
differential sand transports between Scenarios 1 and 2, which represent the sand transport deficits caused by mud
mantling. The differential sand transports were dominated by suspended load, being two orders of magnitude
greater than bedload. During peak periods there were only seaward (positive) differential transports, while during
non‐peak periods there were seaward and landward cycles of small differential transports. Shown in Table 3 are
the cumulative sand transports during typhoon periods at three cross‐sections. On average, suspended load
accounted for 97% of the total load (for either Scenario), the suspended load deficit accounted for >99% of the

Table 3
Cumulative Sand Transports (Suspended Load, Bedload, and Total Load) at Three Cross‐Sections During Typhoon Periods Resulting From Scenarios 1 and 2 (With and
Without Reservoir Mud Releases)
Cumulative sand transport (Mt) Differential sand transport (Mt)
Cross‐section Scenario Suspended load Bedload Total load Suspended load deficit Bedload deficit Total load deficit
River Mouth Scenario 1 2.442 0.081 2.523 0.411 0.002 0.413
Scenario 2 2.853 0.084 2.937
Guandu Bridge Scenario 1 3.489 0.090 3.579 0.580 0.002 0.583
Scenario 2 4.070 0.092 4.162
Taipei Bridge Scenario 1 2.798 0.084 2.882 0.595 0.005 0.600
Scenario 2 3.393 0.089 3.482
Note. Differential sand transports between Scenarios 1 and 2 represent sand transport deficits caused by mud mantling.

HSUEH ET AL. 16 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

Figure 10. Simulation results of Scenario 3 (tracking released muds): bed elevation changes and budgets of reservoir‐released
muds (mud_Res) in four subareas (Dahan River segment, Taipei Bridge segment, Guandu Bridge segment, and mouth zone)
and Taipei Port, after (a) Stage 1, and (b) Stage 2. Summarized in the inset tables are the proportions of mud_Res deposited in
four subareas, exported to offshore and transported to the Taipei Port.

total load deficit. Overall, with reservoir‐released muds and the consequential effect of mud mantling, the mean
sand transport during typhoons was 15% less than that without reservoir‐released muds.
Here is the answer to the first research question posed. Reservoir mud releasing did bear impacts on estuarine and
coastal sediment budgets. The effects of the reservoir‐released muds were indirect, not through direct contri-
butions to the sediment budgets or morphological changes, yet by deposition of muds along the way and mantling
the bed to restrict sand delivery, thus reducing the sand budgets. To what degree were the estuarine and coastal
sand budgets impacted by the reservoir‐released muds then depends on the fate of these muds, as further
elucidated below.

3.3. Scenario 3: Tracking Reservoir‐Released Muds

To track the spatiotemporal distributions of the reservoir‐released muds, the mainstem Tamsui River study area
was split into four subareas (Figure 10), that is, Dahan River segment, Taipei Bridge segment, Guandu Bridge
segment, and mouth zone. The coarse and fine fractions of the reservoir‐released muds are denoted as mud_Res_c
and mud_Res_f, and their sum is denoted as mud_Res.
Distributions of mud_Res after Stages 1 and 2 are shown in Figures 10a and 10b, where the fractional bed
elevation changes and zonal budgets are presented. During Stage 1, a total of 3.48 M ton of muds were released
from the reservoir, 74% of which was exported to offshore, 25% was deposited in the study area, and 1% was
transported to the Taipei Port (see Figure 10a and inset table). Mud deposits declined seaward, from 344 K ton in
the Dahan River segment to 68 K ton in the mouth zone. As noted earlier, the reservoir‐released muds consisted of
40% coarse muds and 60% fine muds, 57% of the coarse muds were deposited while 95% of the fine muds were
exported to offshore (see Figures S15a and S15b in Supporting Information S1 and inset tables). During Stage 2,
those muds deposited during Stage 1 were resuspended and transported by asymmetric tidal currents, resulting in
seaward residual transports (Figures S15c and S15d in Supporting Information S1). After Stage 2, the mud_Res
deposits decreased by 4% while the mud_Res exports increased by 4% (see Figure 10b and inset table). This tidal
flushing took place primarily in the main channels.
Figure 11 shows the time series zonal budgets of mud_Res during Stages 1 and 2. Three results deserve mention
here. First, the budgets of coarse and fine muds during Stage 1 are shown in Figures 11a and 11b. The coarse

HSUEH ET AL. 17 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

Figure 11. Time series budgets of reservoir‐released muds in four subareas (Dahan River segment, Taipei Bridge segment,
Guandu Bridge segment, mouth zone) (a–c) during Stage 1, coarse and fine mud fractions (mud_Res_c and mud_Res_f) and
total muds (mud_Res); (d) during Stage 2, mud_Res.

muds, with a greater ws, were less susceptible to tidal actions, showing increasing trends with cumulative mud
releases. Coarse mud deposits also exhibited a strong trend of seaward decreasing. By contrast, fine mud deposits
varied more closely with the tidal cycle, exhibiting no apparent spatial trend, and was an order of magnitude
smaller than the coarse mud deposits. Second, summing the coarse and fine muds results in Figure 11c, which

HSUEH ET AL. 18 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

exhibits the effects of both cumulative mud deposits and tidal resuspension during Stage 1. Third, the time series
of mud_Res budgets during Stage 2 are shown in Figure 11d, where tidal flushing of mud deposits from the
estuarine segments is clear. This flushing effect exhibits a landward decreasing trend due to attenuation of tidal
wave. After 10 cycles of morphological tide, the mud_Res budget in the wave‐affected mouth zone approached a
dynamic equilibrium, while the mud_Res budgets in the tide‐dominated estuarine segments remained a mild
decreasing trend.
We are now able to answer the second research question posed. During typhoon periods, 74% of the reservoir‐
released muds were exported to offshore, the remaining 25% were deposited along the way. It was these mud
deposits that covered the bed, reduced the sand availability, and caused a 15% reduction in sand transports
compared to a hypothetical scenario of releasing clear‐water floods. The major ingredient of the mud covers was
mud_Res_c, accounting for >90% of the mud_Res deposits. The amount of mud_Res exported to offshore
constituted 62% of the total export, the latter included preexisting sands and muds.

4. Discussion
4.1. Broad Implications
Our results revealed that the synchronized release of mud and flood pulse from the reservoir can be an efficient
measure to mitigate reservoir siltation yet may be a suboptimal solution to addressing coastal sediment deficit. For
the case studied here, during three typhoons in 2008 a total of 11.8 Mm3 muds were released, removing
equivalently >10% of the siltation volume. A majority part (74%) of the reservoir‐released muds were transported
through the estuary and river mouth, and exported to offshore. The remaining part (25%) of the reservoir‐released
muds were deposited along the way, increasing the bed mud contents by 5%–12% compared to a hypothetical
scenario of clear‐water flood releases. On average, mud mantling caused a 15% deficit in fluvial sand transports,
thus reducing the sand budget of river mouth by 14%. The decrease of sand budget suboptimized the efforts to
address coastal sediment deficits, given that these flood‐driven, tributary‐sourced sands play a pivotal role in delta
nourishment, contributing >98% of the sediment budget in river mouth.
From the perspective of reservoir trapping, consider a hypothetical pre‐damming scenario, assuming that
everything else remains as the present. As mentioned in Section 3.1.3, Xindian River dominated tributary sand
supplies, which were mainly sourced from hillslope and channel erosion. Sand supplies from upstream of the
Shihmen Reservoir rarely exceeded 5% of the sediment inflows (NRWRO, 2021), thus the amount of sand trapped
by the reservoir may be neglected. Without a dam the turbidity current inflows during typhoons would be fully
delivered to downstream. This would increase mud deliveries by at least a factor of 3, as the sediment sluicing
ratio of Shihmen Reservoir is at maximum 30% (Section 2.1). With more mud deliveries and deposits, the effect
of mud mantling would be stronger, reducing even more fluvial sand transports and coastal sand budgets. Thus, if
the mud‐release scenario is compared to the pre‐damming scenario, one could argue that the presence of a dam
may bear some merits over demerits. This advantage, however, comes at the cost of trapping at least 70% of
sediment inflows, just like the hypothetical optimal scenario of clear‐water releases that trapped 100% of sedi-
ment inflows. Based on the speculation from the pre‐damming scenario, we can conclude that reservoir desiltation
and delta nourishment are, among others, competing objectives to be compromised in the framework of multi-
objective optimization. The clear‐water release scenario (trapping 100% of sediment inflows) and pre‐damming
scenario (trapping 0% of sediment inflows) are two extreme cases of this framework. For reservoirs with
considerable upstream sand supplies, however, dams could interrupt the natural transport of sands, which would
be then made even worse by mud mantling. Our findings thus highlight that sustainable sediment management
needs to look beyond just a bulk amount of sediment, but for achieving the desired morphologic goals it is critical
to consider how different sediment fractions interact and how they are impacted by human activities.
At the global scale, among the world's 56 cases of reservoir sediment release shown in Figure 2, there are 34% (19/
56) experiencing coastal erosions at the downstream ends (Table S1 in Supporting Information S1). This number
is consistent with those reported by other researchers. For example, Luijendijk et al. (2018), based on analyses of
the satellite derived shorelines, estimated that 24% of the world's sandy coasts are being persistently eroded due in
part to the diminishing sand supplies; Besset et al. (2019), based on data collected from literature and satellite
images for the world's 54 major deltas, found that over a 30‐year period (1985–2015), 54% (29/54) of them were
in net erosion. Further, Table S1 in Supporting Information S1 reveals that among the 19 cases with coastal
erosion, 84% (16/19) had muds as the major type of sediment released, of which 63% (10/16) had reservoir

HSUEH ET AL. 19 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

Figure 12. (a) Event‐wise zonal budgets of reservoir‐released muds (mud_Res) in four subareas. Event‐wise total mud
deposits, and ratios of single‐event mud deposit to mud release (D/R) are also shown. (b) Relation between relative sand
transport deficit (TDef/T0) and degree of bed mud saturation (1–D/R), in which TDef is single‐event sand transport deficit (sum
of suspended‐load and bedload deficits shown in Figures 9j–9k), and T0 is single‐event sand transport of Scenario 2
(hypothetical clear‐water release scenario).

capacities on the order of 106 m3, close to the overall percentage, 75% (42/56), of reservoirs belonging to that
class. Given these resemblances, our study case should bear sufficient generality to offer useful implications for
global reservoirs facing similar problems of siltation and coastal sediment deficit.

4.2. Potential Application

In an attempt to develop optimal strategies for reservoir mud releasing, it may be useful to learn from the evo-
lutions of mud deposit and sand transport with the proceed of mud release events. Figure 12a shows the event‐
wise zonal budgets of mud_Res, which exhibit a seaward decreasing trend. The event‐wise mud releases were
210, 2,538, and 735 k ton, or expressed as the proportions to their sum, 6%: 73%: 21%. The event‐wise mud
deposits were 100, 704, and 73 k ton, or 11% : 80% : 8%. The ratios of the latter to the former are 1.9:1.1:0.4,
implying that for the first event about twice the proportion released was deposited, whereas for the last event less
than half the proportion released was deposited. This is echoed by the ratio of single‐event mud deposit to release,
denoted as D/R (see Figure 12a), which decreased from 48% to 28%–10% with the proceed of mud release events.
The decreasing trend of D/R is necessarily a complex function of the operation and state variables, such as the
mode of mud release, flow magnitude and duration, mud concentration, composition, and settling velocity, flow
velocity, bed shear stress, and most notably the preexisting bed mud content. As shown in Figures 9g–9i, when the
bed mud contents were <0.1 in the pre‐peak period of Sinlaku, bed mud contents remained fixed even though the

HSUEH ET AL. 20 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

flow velocities increased. By contrast, when the bed mud contents were >0.1 in the pre‐peak period of Jangmi,
mud contents decreased as the flow velocities increased. Accordingly, for Jangmi, flushing of the already mud‐
rich bed during the pre‐peak and peak periods, and limited mud deposits during the post‐peak period, were both
responsible for the smallest D/R ratio among the three events.

The single‐event deposit to release ratio D/R is an indicator of the intensity of mud deposition, ranging from 0 (for
zero deposition on a mud‐rich bed) to 1 (for full deposition on a mud‐free bed), hence 1–D/R may be used to
indicate the degree of bed mud saturation. Figure 12b shows the relation between relative sand transport deficit
TDef/T0 and 1–D/R, where TDef/T0 is the ratio of single‐event sand transport deficit to the corresponding sand
transport of the clear‐water flood release scenario. The relative sand transport deficit increases linearly with the
degree of mud saturation (R2 = 0.984), implying that the relative sand transport deficit could be reduced were the
mud release implemented at a smaller value of bed mud content. The trend line also implies that no sand transport
deficit would be present when the degree of bed mud saturation is <0.5.
A potential application of Figure 12b could be made in conjunction with the two‐phase flushing strategy, where a
clear‐water releasing phase is implemented first, followed by a mud releasing phase. The first phase is aimed to
lower the bed mud content, so that sand transport deficit during the second phase can be reduced. As the D/R ratio
of the second phase is not known in advance, the target degree of bed mud saturation 1–D/R in the first phase can
be determined for an established target of TDef/T0. Once 1–D/R (or D/R) is determined, the maximum allowable
mud deposit corresponding to the single‐event mud release can be evaluated, where the mud release is estimated
from the projected mud inflow and the sluicing ratio of the reservoir. The remaining task, hence, is to design a
flood release hydrograph that achieves the goal to constrain mud deposits in the allowable range. A similar two‐
phase strategy has been adopted by the Xiaolangdi Reservoir for flushing the reservoir‐trapped mud deposits
(Wang et al., 2017; X. Wu et al., 2020). How to optimize the two‐phase strategy in order to minimize mud trapped
in the reservoir and mud deposited in the downstream channel, and maximize sand supply to coastal sediment
budget, applying the approach presented here, is a prospective topic for future studies.

4.3. Future Research

Several factors not considered in the present study warrant incorporation in future research. (a) Density currents,
caused by salinity intrusion and SSC stratification during low flows, would confine sediment transport to the near‐
bottom range and deposition to a shorter distance (Wang et al., 2010; G. Wu et al., 2023). Salinity intrusion also
induces baroclinic flows promoting sediment import to the estuary, as well as flocculation leading to greater
settling velocity and mud deposits (Olabarrieta et al., 2018; Zhou et al., 2020). Estuarine baroclinic circulation
would further trap suspended sediment in the topographic low, leading to a local turbidity maximum present
during low flows. To resolve the density stratification effects would require the model to be run in 3D mode. (b)
Sand‐mud interaction, which affects sand erosion flux through the variations of erodibility (critical shear stress for
erosion) and bed roughness (bed shear stress due to skin drag) as a function of bed mud content (Alonso
et al., 2023; Braat et al., 2017). Sand erosion flux would decrease with the increase of bed mud content because of
the reduced erodibility and bed shear stress, not just because of the reduced availability. (c) Decline of sand supply
due to armoring, or supply‐limited condition, in channels downstream of the dam (Wang et al., 2017; X. Wu
et al., 2020) was not considered in our event‐scale morphodynamic study. This factor would potentially reduce
coastal sand budget, thus may be taken into account to assess the evolving sediment budget due to subdecadal‐ to
decadal‐scale regular flood releases. (d) Despite that the effect of storm surge (the rise of sea water level above
the astronomical tide level due to typhoon‐related strong wind, low atmospheric pressure, and wave setup) was
incorporated in our large‐domain estuary model, the large‐scale storm surge effect (with the large‐domain estuary
model covering the full path of a typhoon) was not considered. This large‐scale storm surge effect has the po-
tential to further increase the landward sediment flux (Zhang et al., 2004; Zhu & Wiberg, 2022). To include this
large‐scale effect would require the large‐domain estuary model to be expanded to cover, at least, the entire
Taiwan Strait.

5. Conclusions
We show that the event‐based mud releasing can be effective in mitigating reservoir siltation yet may be a
suboptimal strategy for alleviating coastal sediment deficit. Despite a vast amount of muds were released
from the reservoir during typhoon events, a majority part of these muds were delivered through the estuary

HSUEH ET AL. 21 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

and exported to offshore. Rather, the flood‐driven sands, sourced from tributaries downstream of the dam,
were the major contributor to coastal sediment budget. However, delivery of these sands was restricted by the
remaining part of muds that were deposited along the way, mantling the sand deposits and reducing the sands
available for transport. The sand transport deficit, relative to a hypothetical scenario of releasing clear‐water
floods, increases linearly with the degree of bed mud saturation, a novel finding of this study applicable to
the design of optimal release strategy. Although sand supplies from upstream of the reservoir are minimal in
our case, for reservoirs with considerable upstream sand supplies, however, the dam could interrupt the
natural transport of sands, which is then made even worse by mud mantling. Given broad relevance to global
reservoirs that face the problems of siltation and coastal sediment deficit, our results present insightful im-
plications for optimal release strategy aimed to minimize muds trapped in the reservoir and deposited in the
downstream channel, and maximize sand supplies to coastal sediment budget. Our findings also highlight that
sustainable sediment management needs to look beyond just a bulk amount of sediment, but for accom-
plishing the desired morphologic objectives it is critical to consider how different sediment fractions interact
and how they are impacted by human interventions.

Data Availability Statement

The model input and output files are available at the Zenodo repository (Hsueh & Wu, 2023): https://doi.org/10.
5281/zenodo.8435788. Delft3D source codes are downloadable from the Deltares model repository (https://oss.
deltares.nl/web/delft3d). ERA5 reanalysis wind field and atmospheric pressure field are available from the
ECMWF (https://www.ecmwf.int/). TPXO global tide model is accessible at: https://www.tpxo.net/global.

Acknowledgments References
Research funding was granted by the
National Science and Technology Council Akter, J., Roelvink, D., & van der Wegen, M. (2021). Process‐based modeling deriving a long‐term sediment budget for the Ganges‐Brahmaputra‐
(NSTC) Taiwan to FC Wu (107‐2221‐E‐ Meghna Delta, Bangladesh. Estuarine, Coastal and Shelf Science, 260, 107509. https://doi.org/10.1016/j.ecss.2021.107509
002‐029‐MY3, 109‐2221‐E‐002‐012‐ Allison, M. A., Dallon Weathers, H., & Meselhe, E. A. (2017). Bottom morphology in the Song Hau distributary channel, Mekong River delta,
MY3, 112‐2221‐E‐002‐073‐MY3). Data Vietnam. Continental Shelf Research, 147, 51–61. https://doi.org/10.1016/j.csr.2017.05.010
used in this study were supported by the Allison, M. A., & Meselhe, E. A. (2010). The use of large water and sediment diversions in the lower Mississippi River (Louisiana) for coastal
Harbor and Marine Technology Center restoration. Journal of Hydrology, 387(3–4), 346–360. https://doi.org/10.1016/j.jhydrol.2010.04.001
(IOT, MOTC), Tenth River Management Allison, M. A., Nittrouer, C., Ogston, A., Mullarney, J., & Nguyen, T. (2017). Sedimentation and survival of the Mekong Delta: A case study of
Office, Northern Region Water Resources decreased sediment supply and accelerating rates of relative sea level rise. Oceanography, 30(3), 98–109. https://doi.org/10.5670/oceanog.
Office, and Water Resources Planning 2017.318
Institute (WRA, MOEA), and Taiwan Alonso, C., van Maren, D. S., van Weerdenburg, R. J. A., Huismans, Y., & Wang, Z. B. (2023). Morphodynamic modeling of tidal basins: The role
Ocean Data Bank (NSTC). We appreciate of sand‐mud interaction. Journal of Geophysical Research: Earth Surface, 128(9), e2023JF007391. https://doi.org/10.1029/2023JF007391
Rafael J. P. Schmitt and two anonymous Asaeda, T., & Rashid, M. H. (2012). The impacts of sediment released from dams on downstream sediment bar vegetation. Journal of Hydrology,
reviewers for constructive feedback that 430–431, 25–38. https://doi.org/10.1016/j.jhydrol.2012.01.040
helped improve this paper substantially. Barrera Crespo, P. D., Mosselman, E., Giardino, A., Becker, A., Ottevanger, W., Nabi, M., & Arias‐Hidalgo, M. (2019). Sediment budget analysis
of the Guayas River using a process‐based model. Hydrology and Earth System Sciences, 23(6), 2763–2778. https://doi.org/10.5194/hess‐23‐
2763‐2019
Benedet, L., Dobrochinski, J., Walstra, D., Klein, A., & Ranasinghe, R. (2016). A morphological modeling study to compare different methods of
wave climate schematization and evaluate strategies to reduce erosion losses from a beach nourishment project. Coastal Engineering, 112,
69–86. https://doi.org/10.1016/j.coastaleng.2016.02.005
Besset, M., Anthony, E. J., & Bouchette, F. (2019). Multi‐decadal variations in delta shorelines and their relationship to river sediment supply: An
assessment and review. Earth‐Science Reviews, 193, 199–219. https://doi.org/10.1016/j.earscirev.2019.04.018
Besset, M., Anthony, E. J., & Sabatier, F. (2017). River delta shoreline reworking and erosion in the Mediterranean and Black Seas: The potential
roles of fluvial sediment starvation and other factors. Elementa: Science of the Anthropocene, 5, 54. https://doi.org/10.1525/elementa.139
Blum, M. D., & Roberts, H. H. (2009). Drowning of the Mississippi Delta due to insufficient sediment supply and global sea‐level rise. Nature
Geoscience, 2(7), 488–491. https://doi.org/10.1038/ngeo553
Boateng, I., Bray, M., & Hooke, J. (2012). Estimating the fluvial sediment input to the coastal sediment budget: A case study of Ghana. Geo-
morphology, 138(1), 100–110. https://doi.org/10.1016/j.geomorph.2011.08.028
Braat, L., Pierik, H. J., van Dijk, W. M., van de Lageweg, W. I., Brückner, M. Z. M., van der Meulen, B., & Kleinhans, M. G. (2023). Observed and
modelled tidal bar sedimentology reveals preservation bias against mud in estuarine stratigraphy. The Depositional Record, 9(2), 380–402.
https://doi.org/10.1002/dep2.190
Braat, L., van Kessel, T., Leuven, J. R. F. W., & Kleinhans, M. G. (2017). Effects of mud supply on large‐scale estuary morphology and
development over centuries to millennia. Earth Surface Dynamics, 5(4), 617–652. https://doi.org/10.5194/esurf‐5‐617‐2017
Brandt, S. A., & Swenning, J. (1999). Sedimentological and geomorphological effects of reservoir flushing: The Cachí Reservoir, Costa Rica,
1996. Geografiska Annaler ‐ Series A: Physical Geography, 81(3), 391–407. https://doi.org/10.1111/1468‐0459.00069
Chamoun, S., De Cesare, G., & Schleiss, A. J. (2016). Managing reservoir sedimentation by venting turbidity currents: A review. International
Journal of Sediment Research, 31(3), 195–204. https://doi.org/10.1016/j.ijsrc.2016.06.001
Chen, C.‐N., & Tsai, C.‐H. (2017). Estimating sediment flushing efficiency of a shaft spillway pipe and bed evolution in a reservoir. Water, 9(12),
924. https://doi.org/10.3390/w9120924
Chen, W.‐B., Liu, W. C., Hsu, M. H., & Hwang, C. C. (2015). Modeling investigation of suspended sediment transport in a tidal estuary using a
three‐dimensional model. Applied Mathematical Modelling, 39(9), 2570–2586. https://doi.org/10.1016/j.apm.2014.11.006

HSUEH ET AL. 22 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

Cox, J. R., Huismans, Y., Knaake, S. M., Leuven, J. R. F. W., Vellinga, N. E., van der Vegt, M., et al. (2021). Anthropogenic effects on the
contemporary sediment budget of the lower Rhine‐Meuse Delta channel network. Earth's Future, 9(7), e2020EF001869. https://doi.org/10.
1029/2020ef001869
Darby, S. E., Appeaning Addo, K., Hazra, S., Rahman, M. M., & Nicholls, R. (2020). Fluvial sediment supply and relative sea‐level rise. In R. J.
Nicholls (Ed.), Deltas in the Anthropocene (pp. 103–126). Palgrave Macmillan.
Deltares. (2014a). Delft3D‐FLOW, user manual. Version 3.15. Delft.
Deltares. (2014b). Delft3D‐WAVE, user manual. Version 3.05. Delft.
Dethier, E. N., Renshaw, C. E., & Magilligan, F. J. (2022). Rapid changes to global river suspended sediment flux by humans. Science, 376(6600),
1447–1452. https://doi.org/10.1126/science.abn7980
Dissanayake, D. M. P. K., Wurpts, A., Miani, M., Knaack, H., Niemeyer, H., & Roelvink, J. (2012). Modelling morphodynamic response of a tidal
basin to an anthropogenic effect: Ley Bay, East Frisian Wadden Sea – Applying tidal forcing only and different sediment fractions. Coastal
Engineering, 67, 14–28. https://doi.org/10.1016/j.coastaleng.2012.04.001
Dunn, F. E., Darby, S. E., Nicholls, R. J., Cohen, S., Zarfl, C., & Fekete, B. M. (2019). Projections of declining fluvial sediment delivery to major
deltas worldwide in response to climate change and anthropogenic stress. Environmental Research Letters, 14(8), 084034. https://doi.org/10.
1088/1748‐9326/ab304e
Egbert, G. D., & Erofeeva, S. Y. (2002). Efficient inverse modeling of barotropic ocean tides. Journal of Atmospheric and Oceanic Technology,
19(2), 183–204. https://doi.org/10.1175/1520‐0426(2002)019<0183:eimobo>2.0.co;2
Espa, P., Batalla, R. J., Brignoli, M. L., Crosa, G., Gentili, G., & Quadroni, S. (2019). Tackling reservoir siltation by controlled sediment flushing:
Impact on downstream fauna and related management issues. PLoS One, 14(6), e0218822. https://doi.org/10.1371/journal.pone.0218822
Fredsøe, J. (1984). Turbulent boundary layer in wave‐current motion. Journal of Hydraulic Engineering, 110(8), 1103–1120. https://doi.org/10.
1061/(asce)0733‐9429(1984)110:8(1103)
Fruchard, F., & Camenen, B. (2012). Reservoir sedimentation: Different type of flushing—Friendly flushing example of genissiat dam flushing.
In Proceedings of the International Symposium on Dams for a Changing World. ICOLD.
Geleynse, N., Storms, J. E. A., Stive, M. J. F., Jagers, H. R. A., & Walstra, D. J. R. (2010). Modeling of a mixed‐load fluvio‐deltaic system.
Geophysical Research Letters, 37(5), L05402. https://doi.org/10.1029/2009GL042000
Gelfenbaum, G., Stevens, A. W., Miller, I., Warrick, J. A., Ogston, A. S., & Eidam, E. (2015). Large‐scale dam removal on the Elwha River,
Washington, USA: Coastal geomorphic change. Geomorphology, 246, 649–668. https://doi.org/10.1016/j.geomorph.2015.01.002
Guo, L., van der Wegen, M., Roelvink, J. A., & He, Q. (2014). The role of river flow and tidal asymmetry on 1‐D estuarine morphodynamics.
Journal of Geophysical Research: Earth Surface, 119(11), 2315–2334. https://doi.org/10.1002/2014jf003110
Hayes, M. O. (1979). Barrier island morphology as a function of wave and tide regime. In S. Leatherman (Ed.), Barrier Islands from the Gulf of St.
Lawrence to the Gulf of Mexico (pp. 1–29). Academic Press.
Hibma, A., de Vriend, H., & Stive, M. (2003). Numerical modelling of shoal pattern formation in well‐mixed elongated estuaries. Estuarine,
Coastal and Shelf Science, 57(5–6), 981–991. https://doi.org/10.1016/s0272‐7714(03)00004‐0
Hong, E., Ou, C. H., Tseng, S. S., Kuo, N. J., Huang, T. C., & Lee, W. C. (2000). The future impact of the Tanshui Harbor on its nearby coastal
environment. Journal of the Geological Society of China, 43, 341–360.
Hopkins, J., Elgar, S., & Raubenheimer, B. (2018). Storm impact on morphological evolution of a sandy inlet. Journal of Geophysical Research:
Oceans, 123(8), 5751–5762. https://doi.org/10.1029/2017JC013708
Hsieh, T.‐C., Ding, Y., Yeh, K. C., & Jhong, R. K. (2020). Investigation of morphological changes in the Tamsui River estuary using an integrated
coastal and estuarine processes model. Water, 12(4), 1084. https://doi.org/10.3390/w12041084
Hsu, M.‐H., Wu, C. R., Liu, W. C., & Kuo, A. Y. (2006). Investigation of turbidity maximum in a mesotidal estuary, Taiwan. Journal of the
American Water Resources Association, 42(4), 901–914. https://doi.org/10.1111/j.1752‐1688.2006.tb04503.x
Hsu, S. T., & Wu, C.‐H. (2019). The conversion of Shihmen Reservoir’s #2 penstock into a silt‐sluice way. In Proceedings of the 3rd International
Workshop on Sediment Bypass Tunnels. NTU.
Hsueh, Y.‐T., & Wu, F.‐C. (2023). Dataset accompanying the publication: Reservoir mud releasing may suboptimize fluvial sand supply to coastal
sediment budget: Modeling the impact of Shihmen Reservoir case on Tamsui River estuary [Dataset]. Zenodo. https://doi.org/10.5281/zenodo.
8435788
Hu, K., Ding, P., Wang, Z., & Yang, S. (2009). A 2D/3D hydrodynamic and sediment transport model for the Yangtze Estuary, China. Journal of
Marine Systems, 77(1–2), 114–136. https://doi.org/10.1016/j.jmarsys.2008.11.014
Kemp, G. P., Day, J. W., Rogers, J. D., Giosan, L., & Peyronnin, N. (2016). Enhancing mud supply from the Lower Missouri River to the
Mississippi River Delta USA: Dam bypassing and coastal restoration. Estuarine, Coastal and Shelf Science, 183, 304–313. https://doi.org/10.
1016/j.ecss.2016.07.008
Kondolf, G. M., Gao, Y., Annandale, G. W., Morris, G. L., Jiang, E., Zhang, J., et al. (2014). Sustainable sediment management in reservoirs and
regulated rivers: Experiences from five continents. Earth's Future, 2(5), 256–280. https://doi.org/10.1002/2013ef000184
Kondolf, G. M., Schmitt, R. J., Carling, P., Darby, S., Arias, M., Bizzi, S., et al. (2018). Changing sediment budget of the Mekong: Cumulative
threats and management strategies for a large river basin. Science of the Total Environment, 625, 114–134. https://doi.org/10.1016/j.scitotenv.
2017.11.361
Kondolf, G. M., & Yi, J. (2022). Dam renovation to prolong reservoir life and mitigate dam impacts. Water, 14(9), 1464. https://doi.org/10.3390/
w14091464
Lai, S. Y. J., & Wu, F.‐C. (2021). Two‐stage transition from Gilbert to hyperpycnal delta in reservoir. Geophysical Research Letters, 48(14),
e2021GL093661. https://doi.org/10.1029/2021GL093661
Lee, F.‐Z., Lai, J. S., & Sumi, T. (2022). Reservoir sediment management and downstream river impacts for sustainable water resources—Case
study of Shihmen Reservoir. Water, 14(3), 479. https://doi.org/10.3390/w14030479
Lesser, G. R. (2009). An approach to medium‐term coastal morphological modelling (PhD thesis). Delft University of Technology and
UNESCO‐IHE.
Lesser, G. R., Roelvink, J., van Kester, J., & Stelling, G. (2004). Development and validation of a three‐dimensional morphological model.
Coastal Engineering, 51(8–9), 883–915. https://doi.org/10.1016/j.coastaleng.2004.07.014
Lin, P.‐C., Chou, H.‐T., Capart, H., Lin, W.‐J., & Ho, L.‐S. (2007). Sediment transport at Danshui River Mouth and surrounding coastlines.
Report 96‐58‐7272. Harbor and Marine Technology Center, Institute of Transportation.
Lin, P.‐C., Wei, C.‐H., & Ho, L.‐S. (2017). 2016 field Investigation of sediment transport in the shore vicinity of harbors. Report 106‐036‐7936.
Harbor and Marine Technology Center, Institute of Transportation.
Liu, W.‐C., Chen, W. B., Cheng, R. T., Hsu, M. H., & Kuo, A. Y. (2007). Modeling the influence of river discharge on salt intrusion and residual
circulation in Danshuei River estuary, Taiwan. Continental Shelf Research, 27(7), 900–921. https://doi.org/10.1016/j.csr.2006.12.005

HSUEH ET AL. 23 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

Liu, W.‐C., Liu, H. M., & Huang, W. C. (2022). Flood‐ebb and discharge variations in observed salinity and suspended sediment in a mesotidal
estuary. The Standard, 2, 209–225. https://doi.org/10.3390/standards2020016
Luan, H. L., Ding, P. X., Wang, Z. B., & Ge, J. Z. (2017). Process‐based morphodynamic modeling of the Yangtze Estuary at a decadal timescale:
Controls on estuarine evolution and future trends. Geomorphology, 290, 347–364. https://doi.org/10.1016/j.geomorph.2017.04.016
Luijendijk, A., Hagenaars, G., Ranasinghe, R., Baart, F., Donchyts, G., & Aarninkhof, S. (2018). The state of the world’s beaches. Scientific
Reports, 8(1), 6641. https://doi.org/10.1038/s41598‐018‐24630‐6
Luijendijk, A. P., Ranasinghe, R., de Schipper, M. A., Huisman, B. A., Swinkels, C. M., Walstra, D. J., & Stive, M. J. (2017). The initial
morphological response of the sand engine: A process‐based modelling study. Coastal Engineering, 119, 1–14. https://doi.org/10.1016/j.
coastaleng.2016.09.005
Merwade, V. M., Maidment, D. R., & Goff, J. A. (2006). Anisotropic considerations while interpolating river channel bathymetry. Journal of
Hydrology, 331(3–4), 731–741. https://doi.org/10.1016/j.jhydrol.2006.06.018
Meselhe, E. A., Sadid, K. M., & Allison, M. A. (2016). Riverside morphological response to pulsed sediment diversions. Geomorphology, 270,
184–202. https://doi.org/10.1016/j.geomorph.2016.07.023
Morris, G. L. (2020a). Classification of management alternatives to combat reservoir sedimentation. Water, 12(3), 861. https://doi.org/10.3390/
w12030861
Morris, G. L., & Fan, J. (1998). Reservoir sedimentation handbook. McGraw‐Hill.
Nardin, W., Lera, S., & Nienhuis, J. (2020). Effect of offshore waves and vegetation on the sediment budget in the Virginia Coast Reserve (VA).
Earth Surface Processes and Landforms, 45(12), 3055–3068. https://doi.org/10.1002/esp.4951
Nittrouer, J. A., & Viparelli, E. (2014). Sand as a stable and sustainable resource for nourishing the Mississippi River delta. Nature Geoscience,
7(5), 350–354. https://doi.org/10.1038/ngeo2142
NRWRO. (2021). Sediment transport monitoring facility maintenance and flood event observations in Shihmen Reservoir (1/2). Report
MOEAWRA1100123. Northern Region Water Resources Office, WRA, MOEA.
NRWRO. (2022). Shihmen reservoir website. Northern Region Water Resources Office, WRA, MOEA. Retrieved from https://shihmen.wranb.
gov.tw/
Olabarrieta, M., Geyer, W. R., Coco, G., Friedrichs, C. T., & Cao, Z. (2018). Effects of density‐driven flows on the long‐term morphodynamic
evolution of funnel‐shaped estuaries. Journal of Geophysical Research: Earth Surface, 123(11), 2901–2924. https://doi.org/10.1029/
2017jf004527
Partheniades, E. (1965). Erosion and deposition of cohesive soils. Journal of the Hydraulics Division, 91(1), 105–139. https://doi.org/10.1061/
jyceaj.0001165
Roelvink, D., & Reniers, A. (2012). A guide to modeling coastal morphology. World Scientific.
Roelvink, D., & Walstra, D.‐J. (2004). Keeping it simple by using complex models. In M. S. Altinakar (Ed.), Advances in Hydro‐science and
Engineering (Vol. 6, pp. 335–346). National Center for Computational Hydroscience and Engineering.
Rovira, A., & Ibàñez, C. (2007). Sediment management options for the lower Ebro River and its delta. Journal of Soils and Sediments, 7(5),
285–295. https://doi.org/10.1065/jss2007.08.244
Schleiss, A. J., Franca, M. J., Juez, C., & De Cesare, G. (2016). Reservoir sedimentation. Journal of Hydraulic Research, 54(6), 595–614. https://
doi.org/10.1080/00221686.2016.1225320
Schmitt, R. J. P., & Minderhoud, P. S. J. (2023). Data, knowledge, and modeling challenges for science‐informed management of river deltas. One
Earth, 6(3), 216–235. https://doi.org/10.1016/j.oneear.2023.02.010
Schmitt, R. J. P., Rubin, Z., & Kondolf, G. (2017). Losing ground ‐ Scenarios of land loss as consequence of shifting sediment budgets in the
Mekong Delta. Geomorphology, 294, 58–69. https://doi.org/10.1016/j.geomorph.2017.04.029
Soulsby, R. L., Hamm, L., Klopman, G., Myrhaug, D., Simons, R., & Thomas, G. (1993). Wave‐current interaction within and outside the bottom
boundary layer. Coastal Engineering, 21(1–3), 41–69. https://doi.org/10.1016/0378‐3839(93)90045‐a
Stephens, J. D., Allison, M., Di Leonardo, D., Weathers, H., Ogston, A., McLachlan, R., et al. (2017). Sand dynamics in the Mekong River channel
and export to the coastal ocean. Continental Shelf Research, 147, 38–50. https://doi.org/10.1016/j.csr.2017.08.004
Su, C.‐H., Liaw, C.‐T., Huang, M.‐H., Luo, G.‐S., Wei, C.‐H., Lee, J.‐D., et al. (2021). Annual statistic report of oceanographical observation
data in Taipei offshore region at 2019. Institute of Transportation, Ministry of Transportation and Communications.
Sumi, T. (2008). Evaluation of efficiency of reservoir sediment flushing in Kurobe River. In T. Sekiguchi (Ed.), Proceedings 4th International
Conference on scour and erosion (pp. 608–613).
Syvitski, J. P. M., Vörösmarty, 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), 376–380. https://doi.org/10.1126/science.1109454
Toniolo, H., Parker, G., & Voller, V. (2007). Role of ponded turbidity currents in reservoir trap efficiency. Journal of Hydraulic Engineering,
133(6), 579–595. https://doi.org/10.1061/(asce)0733‐9429(2007)133:6(579)
TRMO. (2015). Impact assessment of sediment transport at Tansui estuary on surrounding coast (2/2). Tenth River Management Office, WRA,
MOEA.
Tu, L. X., Thanh, V. Q., Reyns, J., Van, S. P., Anh, D. T., Dang, T. D., & Roelvink, D. (2019). Sediment transport and morphodynamical modeling
on the estuaries and coastal zone of the Vietnamese Mekong Delta. Continental Shelf Research, 186, 64–76. https://doi.org/10.1016/j.csr.2019.
07.015
Valencia, A. A., Storms, J. E. A., Walstra, D. R., van der Vegt, H., & Jagers, H. R. A. (2023). The impact of clastic syn‐sedimentary compaction on
fluvial‐dominated delta morphodynamics. The Depositional Record, 9(2), 233–252. https://doi.org/10.1002/dep2.219
van der Vegt, H., Storms, J., Walstra, D., & Howes, N. (2016). Can bed load transport drive varying depositional behaviour in river delta en-
vironments? Sedimentary Geology, 345, 19–32. https://doi.org/10.1016/j.sedgeo.2016.08.009
van der Vegt, H., Storms, J. E., Walstra, D. R., Nordahl, K., Howes, N. C., & Martinius, A. W. (2020). Grain size fractionation by process‐driven
sorting in sandy to muddy deltas. The Depositional Record, 6(1), 217–235. https://doi.org/10.1002/dep2.85
van der Wegen, M., Dastgheib, A., Jaffe, B. E., & Roelvink, D. (2011). Bed composition generation for morphodynamic modeling: Case study of
San Pablo Bay in California, USA. Ocean Dynamics, 61(2–3), 173–186. https://doi.org/10.1007/s10236‐010‐0314‐2
van der Wegen, M., Jaffe, B. E., & Roelvink, J. A. (2011). Process‐based, morphodynamic hindcast of decadal deposition patterns in San Pablo
Bay, California, 1856–1887. Journal of Geophysical Research, 116(F2), F02008. https://doi.org/10.1029/2009jf001614
van der Wegen, M., & Roelvink, J. A. (2008). Long‐term morphodynamic evolution of a tidal embayment using a two‐dimensional, process‐based
model. Journal of Geophysical Research, 113(C3), C03016. https://doi.org/10.1029/2006jc003983
van der Wegen, M., Wang, Z. B., Savenije, H. H. G., & Roelvink, J. A. (2008). Long‐term morphodynamic evolution and energy dissipation in a
coastal plain, tidal embayment. Journal of Geophysical Research, 113(F3), F03001. https://doi.org/10.1029/2007jf000898
van Rijn, L. C. (1993). Principles of sediment transport in rivers, estuaries and coastal seas. Aqua Publications.

HSUEH ET AL. 24 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

van Rijn, L. C., Roelvink, J. A., & Horst, W. T. (2001). Approximation formulae for sand transport by currents and waves and implementation in
DELFT‐MOR. Hydraulic Engineering Report Z3054.20/40. Deltares (WL). Retrieved from http://resolver.tudelft.nl/uuid:c226bf0a‐79d6‐
4357‐80ff‐c68c4d5f0416
Wang, H., Bi, N., Wang, Y., Saito, Y., & Yang, Z. (2010). Tide‐modulated hyperpycnal flows off the Huanghe (Yellow River) mouth, China.
Earth Surface Processes and Landforms, 35(11), 1315–1329. https://doi.org/10.1002/esp.2032
Wang, H., Wu, X., Bi, N., Li, S., Yuan, P., Wang, A., et al. (2017). Impacts of the dam‐orientated water‐sediment regulation scheme on the lower
reaches and delta of the Yellow River (Huanghe): A review. Global and Planetary Change, 157, 93–113. https://doi.org/10.1016/j.gloplacha.
2017.08.005
Warrick, J. A., Bountry, J. A., East, A. E., Magirl, C. S., Randle, T. J., Gelfenbaum, G., et al. (2015). Large‐scale dam removal on the Elwha River,
Washington, USA: Source‐to‐sink sediment budget and synthesis. Geomorphology, 246, 729–750. https://doi.org/10.1016/j.geomorph.2015.
01.010
Warrick, J. A., Stevens, A. W., Miller, I. M., Harrison, S. R., Ritchie, A. C., & Gelfenbaum, G. (2019). World’s largest dam removal reverses
coastal erosion. Scientific Reports, 9(1), 13968. https://doi.org/10.1038/s41598‐019‐50387‐7
Willmott, C. J. (1981). On the validation of models. Physical Geography, 2, 184–194. https://doi.org/10.1080/02723646.1981.10642213
WRPI. (2008). Observation and measurement of sediment transport and application/development of density current simulation model in Shihmen
Reservoir (1/2). Report 096D27125050000. Water Resources Planning Institute.
Wu, G., Wang, K., Liang, B., Wu, X., Wang, H., Li, H., & Shi, B. (2023). Modeling the morphological responses of the Yellow River Delta to the
water‐sediment regulation scheme: The role of impulsive river floods and density‐driven flows. Water Resources Research, 59(7),
e2022WR033003. https://doi.org/10.1029/2022WR033003
Wu, X., Bi, N., Syvitski, J., Saito, Y., Xu, J., Nittrouer, J. A., et al. (2020). Can reservoir regulation along the Yellow River be a sustainable way to
save a sinking delta? Earth's Future, 8(11), e2020EF001587. https://doi.org/10.1029/2020ef001587
Wu, X., Wang, H., Bi, N., Xu, J., Nittrouer, J. A., Yang, Z., et al. (2021). Impact of artificial floods on the quantity and grain size of river‐borne
sediment: A case study of a dam regulation scheme in the yellow river catchment. Water Resources Research, 57(5), e2021WR029581. https://
doi.org/10.1029/2021WR029581
Yang, S. L., Milliman, J., Xu, K., Deng, B., Zhang, X., & Luo, X. (2014). Downstream sedimentary and geomorphic impacts of the Three Gorges
Dam on the Yangtze River. Earth‐Science Reviews, 138, 469–486. https://doi.org/10.1016/j.earscirev.2014.07.006
Zhang, Q., Lei, Y., & Wang, C. (2004). Analysis of sediment outflow of turbidity current in Naodehai Reservoir In. M. Wieland (Ed.), New
developments in dam engineering (pp. 1053–1057). Taylor and Francis.
Zhou, Z., Chen, L., Tao, J., Gong, Z., Guo, L., van der Wegen, M., et al. (2020). The role of salinity in fluvio‐deltaic morphodynamics: A long‐
term modelling study. Earth Surface Processes and Landforms, 45(3), 590–604. https://doi.org/10.1002/esp.4757
Zhu, Q., Wang, Y. P., Gao, S., Zhang, J., Li, M., Yang, Y., & Gao, J. (2017). Modeling morphological change in anthropogenically controlled
estuaries. Anthropocene, 17, 70–83. https://doi.org/10.1016/j.ancene.2017.03.001
Zhu, Q., & Wiberg, P. L. (2022). The importance of storm surge for sediment delivery to microtidal marshes. Journal of Geophysical Research:
Earth Surface, 127(9), e2022JF006612. https://doi.org/10.1029/2022JF006612

References From the Supporting Information

Adam, E., & Suleiman, M. (2022). Reservoir sediment management practices in Sudan: A case study of Khashm El‐Girba Dam. In C. Sumi (Ed.),
Wadi flash floods, natural disaster science and mitigation engineering: DPRI reports (pp. 455–471). Springer Nature.
Alliau, D., Peteuil, C., & Huybrechts, N. (2016). Using an eco‐friendly flushing event to calibrate 3D sediment transport model through a
reservoir: The case study of Champagneux run‐of‐river dam on the Rhöne River, France. In S. Bourban (Ed.), Proceedings of the 23rd
TELEMAC‐MASCARET User Conference 2016 (pp. 49–53). HR Wallingford.
Avendaño‐Salas, C., Sanz Montero, M. E., & Cobo Rayán, R. (2000). State of the art of reservoir sedimentation management in Spain. In
Proceedings of International Workshop and Symposium on Reservoir Sedimentation Management (pp. 27–34). WEC.
Atkinson, E. (1996). The feasibility of flushing sediment from reservoirs. HR Wallingford.
Besnier, A.‐L., Secher, M., & Tassi, P. (2016). Numerical modelling of the sediment dynamics in the Saint‐Martin‐la‐Porte reservoir, France. In G.
Constantinescu (Ed.), River flow 2016 (pp. 1426–1434).
Booij, N., Ris, R. C., & Holthuijsen, L. H. (1999). A third‐generation wave model for coastal regions: 1. Model description and validation. Journal
of Geophysical Research, 104(C4), 7649–7666. https://doi.org/10.1029/98jc02622
Camenen, B., Naudet, G., Dramais, G., Le Coz, J., & Paquier, A. (2019). A multi‐technique approach for evaluating sand dynamics in a complex
engineered piedmont river system. Science of the Total Environment, 657, 485–497. https://doi.org/10.1016/j.scitotenv.2018.11.394
Cattanéo, F., Guillard, J., Diouf, S., O'Rourke, J., & Grimardias, D. (2021). Mitigation of ecological impacts on fish of large reservoir sediment
management through controlled flushing – The case of the Verbois dam (Rhône River, Switzerland). Science of the Total Environment, 756,
144053. https://doi.org/10.1016/j.scitotenv.2020.144053
Chaudhry, M. A., Habib‐ur‐Rehman, M., Akhtar, M. N., & Hashmi, H. N. (2014). Modeling sediment deposition and sediment flushing through
reservoirs using 1‐D numerical model. Arabian Journal for Science and Engineering, 39(2), 647–658. https://doi.org/10.1007/s13369‐013‐
0670‐6
Chaudhry, M. R. (1982). Flushing operations of Warsak reservoir sediment. In Proceedings of the Pakistan Engineering Congress (Vol. 454, pp.
246–265).
Deltares. (2014c). RGFGRID, user manual. Version 4.00. Delft.
den Boer, V. J. E. (2011). Preliminary study of the flushing operations of the Langmann reservoir, Austria (Master thesis). Delft University of
Technology.
Dhivert, E., Grosbois, C., Coynel, A., Lefèvre, I., & Desmet, M. (2015). Influences of major flood sediment inputs on sedimentary and
geochemical signals archived in a reservoir core (Upper Loire Basin, France). Catena, 126, 75–85. https://doi.org/10.1016/j.catena.2014.
10.030
Doretto, A., Bo, T., Bona, F., Apostolo, M., Bonetto, D., & Fenoglio, S. (2019). Effectiveness of artificial floods for benthic community recovery
after sediment flushing from a dam. Environmental Monitoring and Assessment, 191(2), 88. https://doi.org/10.1007/s10661‐019‐7232‐7
Eriksen, K., & Easthouse, K. (2010). Howard Hanson Dam sediment management project. In Paper presented at 2nd Joint Federal Interagency
Conference. Las Vegas.
Folegot, S., Bruno, M. C., Larsen, S., Kaffas, K., Pisaturo, G. R., Andreoli, A., et al. (2021). The effects of a sediment flushing on Alpine
macroinvertebrate communities. Hydrobiologia, 848(17), 3921–3941. https://doi.org/10.1007/s10750‐021‐04608‐8

HSUEH ET AL. 25 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

Frémion, F., Bordas, F., Mourier, B., Lenain, J. F., Kestens, T., & Courtin‐Nomade, A. (2016a). Influence of dams on sediment continuity: A study
case of a natural metallic contamination. Science of the Total Environment, 547, 282–294. https://doi.org/10.1016/j.scitotenv.2016.01.023
Frémion, F., Courtin‐Nomade, A., Bordas, F., Lenain, J. F., Jugé, P., Kestens, T., & Mourier, B. (2016b). Impact of sediments resuspension on
metal solubilization and water quality during recurrent reservoir sluicing management. Science of the Total Environment, 562, 201–215. https://
doi.org/10.1016/j.scitotenv.2016.03.178
Furukawa, H., Fukushige, Y., Okumura, H., Asaoka, Y., & Nagabayashi, H. (2019). Measurement of non‐uniform suspended sediment transport
during sluicing and flushing operations at the Seto‐Ishi Dam on the Kumagawa River. Journal of Japan Society of Civil Engineers, Ser. B1
(Hydraulic Engineering), 75(2), 853–858. https://doi.org/10.2208/jscejhe.75.2_i_853
Gellis, A. C., Webb, R. M. T., McIntyre, S. C., & Wolfe, W. J. (2006). Land‐use effects on erosion, sediment yields, and reservoir sedimentation:
A case study in the Lago Loíza Basin, Puerto Rico. Physical Geography, 27(1), 39–69. https://doi.org/10.2747/0272‐3646.27.1.39
Gibson, S., & Boyd, P. (2016). Monitoring, measuring, and modeling a reservoir flush on the Niobrara River in the Sandhills of Nebraska. In G.
Constantinescu (Ed.), River flow 2016 (pp. 1448–1455).
Gibson, S., & Crain, J. (2019). Modeling sediment concentrations during a drawdown reservoir flush: Simulating the fall creek operations with
HEC‐RAS. ERDC/TN RSM‐19‐7. U.S. Army Engineer Research and Development Center.
Guertault, L., Camenen, B., Peteuil, C., & Paquier, A. (2014). Long term evolution of a dam reservoir subjected to regular flushing events.
Advances in Geosciences, 39, 89–94. https://doi.org/10.5194/adgeo‐39‐89‐2014
Guertault, L., Camenen, B., Paquier, A., & Peteuil, C. (2018). A one‐dimensional process‐based approach to study reservoir sediment dynamics
during management operations. Earth Surface Processes and Landforms, 43(2), 373–386. https://doi.org/10.1002/esp.4249
Hassanzadeh, Y. (1995). The removal of reservoir sediment. Water International, 20(3), 151–154. https://doi.org/10.1080/02508069508686467
Huang, J., Greimann, B., & Kimbrel, S. (2019). Simulation of sediment flushing in Paonia reservoir of Colorado. Journal of Hydraulic Engi-
neering, 145(12), 06019015. https://doi.org/10.1061/(asce)hy.1943‐7900.0001651
IHA. (2022). Sediment management case studies: India ‐ Nathpa Jhakri. International Hydropower Association. Retrieved from https://www.
hydropower.org/sediment‐management‐case‐studies/india‐nathpa‐jhakri
Kaffas, K., Pisinaras, V., Al Sayah, M. J., Santopietro, S., & Righetti, M. (2021). A USLE‐based model with modified LS‐factor combined with
sediment delivery module for Alpine basins. Catena, 207, 105655. https://doi.org/10.1016/j.catena.2021.105655
Launay, M., Dugué, V., Faure, J. B., Coquery, M., Camenen, B., & Le Coz, J. (2019). Numerical modelling of the suspended particulate matter
dynamics in a regulated river network. Science of the Total Environment, 665, 591–605. https://doi.org/10.1016/j.scitotenv.2019.02.015
Legout, C., Droppo, I., Coutaz, J., Bel, C., & Jodeau, M. (2018). Assessment of erosion and settling properties of fine sediments stored in cobble
bed rivers: The Arc and Isère alpine rivers before and after reservoir flushing. Earth Surface Processes and Landforms, 43(6), 1295–1309.
https://doi.org/10.1002/esp.4314
Leobacher, A., & Blauhut, A. (2010). Gerlos power station/Gmünd dam – Stabilization of a reservoir slope (Grasegger slope). Geomechanics and
Tunnelling, 3(5), 462–469. https://doi.org/10.1002/geot.201000041
Lepage, H., Launay, M., Le Coz, J., Angot, H., Miège, C., Gairoard, S., et al. (2020). Impact of dam flushing operations on sediment dynamics and
quality in the upper Rhône River, France. Journal of Environmental Management, 255, 109886. https://doi.org/10.1016/j.jenvman.2019.
109886
Liu, D., Nie, Q., Xiong, C., Yu, Z., Lv, Y., Zhang, S., et al. (2022). Sediment particle size composition in the riparian zone of the Three Gorges
Reservoir. Frontiers of Environmental Science, 10, 820700. https://doi.org/10.3389/fenvs.2022.820700
Liu, W.‐C., Hsu, M. H., & Kuo, A. Y. (2002). Modelling of hydrodynamics and cohesive sediment transport in Tanshui River estuarine system,
Taiwan. Marine Pollution Bulletin, 44(10), 1076–1088. https://doi.org/10.1016/s0025‐326x(02)00160‐1
Luijendijk, A. P., Schipper, M., & Ranasinghe, R. (2019). Morphodynamic acceleration techniques for multi‐timescale predictions of complex
sandy interventions. Journal of Marine Science and Engineering, 7(3), 78. https://doi.org/10.3390/jmse7030078
Marot, D., Bouchard, J. P., & Alexis, A. (2005). Reservoir bank deformation modeling: Application to Grangent reservoir. Journal of Hydraulic
Engineering, 131(7), 586–595. https://doi.org/10.1061/(asce)0733‐9429(2005)131:7(586)
Meile, T., Bretz, N. V., Imboden, B., & Boillat, J. L. (2014). Reservoir sedimentation management at Gebidem dam (Switzerland). In M. R.
Schleiss (Ed.), Reservoir sedimentation (pp. 245–255). Taylor and Francis.
Milne, B. E. (1969). A note on siltation in the Mangahao power project. In J. A. Hayward (Ed.), Watershed management, Part 2, Proceeedings of a
Symposium on Watershed Management in Water Resources Development (pp. 42–50). New Zealand Agricultural Engineering Institute.
Moridi, A., & Yazdi, J. (2017). Sediment flushing of reservoirs under environmental considerations. Water Resources Management, 31(6),
1899–1914. https://doi.org/10.1007/s11269‐017‐1620‐y
Morris, G. L. (2020b). Solving persistent water supply problems in Puerto Rico: Critical problems and actions. In White paper for ASCE
infrastructure committee. San Juan.
Nukazawa, K., Kajiwara, S., Saito, T., & Suzuki, Y. (2020). Preliminary assessment of the impacts of sediment sluicing events on stream insects in
the Mimi River, Japan. Ecological Engineering, 145, 105726. https://doi.org/10.1016/j.ecoleng.2020.105726
OJSCEPP. (2018). Initial environmental examination, Kyrgyz republic: Uch‐Kurgan hydropower plant modernization project. Document pre-
pared by the consulting joint venture tractebel engineering and endustriel elektrik maden LLC for the Asian development bank. In Open joint
stock company electric power plants. Bishkek.
Peteuil, C., Jodeau, M., De Linares, M., Valette, E., Alliau, D., Wirz, C., et al. (2018). Toward an operational approach for the characterization and
modelling of fine sediments dynamics in reservoirs. In A. Paquier & N. Rivière (Ed.), River flow 2018—9th International Conference on
Fluvial Hydraulics, E3S web of conferences (Vol. 40, p. 03028). https://doi.org/10.1051/e3sconf/20184003028
Ranasinghe, R., Swinkels, C., Luijendijk, A., Roelvink, D., Bosboom, J., Stive, M., & Walstra, D. (2011). Morphodynamic upscaling with the
MORFAC approach: Dependencies and sensitivities. Coastal Engineering, 58(8), 806–811. https://doi.org/10.1016/j.coastaleng.2011.03.010
Reckendorfer, W., Badura, H., & Schütz, C. (2019). Drawdown flushing in a chain of reservoirs ‐ Effects on grayling populations and implications
for sediment management. Ecology and Evolution, 9(3), 1437–1451. https://doi.org/10.1002/ece3.4865
Ren, S., Zhang, B., Wang, W. J., Yuan, Y., & Guo, C. (2021). Sedimentation and its response to management strategies of the Three Gorges
Reservoir, Yangtze River, China. Catena, 199, 105096. https://doi.org/10.1016/j.catena.2020.105096
Roelvink, J. A. (2006). Coastal morphodynamic evolution techniques. Coastal Engineering, 53(2–3), 277–287. https://doi.org/10.1016/j.
coastaleng.2005.10.015
Rollo, C., Bonnani, S., Sayah, S. M., Arbolí, J., & Braghini, M. (2018). Sedimentation management of Cerro del Águila Reservoir in Peru: A
“mobile dam” for a highly erosive Andean watershed. In Proceedings of the 26th international congress on large dams. CRC Press.
Schenk, L. N., & Bragg, H. M. (2014). Assessment of suspended‐sediment transport, bedload, and dissolved oxygen during a short‐term
drawdown of Fall Creek Lake, Oregon, winter 2012–13. In Open file report 2014–1114. U.S. Geological Survey.

HSUEH ET AL. 26 of 27
 19447973, 2024, 6, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023WR036701 by South African Medical Research, Wiley Online Library on [31/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Water Resources Research 10.1029/2023WR036701

Shelley, J., Hotchkiss, R. H., Boyd, P., & Gibson, S. (2022). Discharging sediment downstream: Case studies in cost effective, environmentally
acceptable reservoir sediment management in the United States. Journal of Water Resources Planning and Management, 148(2), 05021028.
https://doi.org/10.1061/(asce)wr.1943‐5452.0001494
Shelley, J. E., Boyd, P. M., Dahl, T. A., Floyd, I. E., & Ramos‐Villanueva, M. (2018). Reservoir sediment management workshop for regulators,
planners, and managers, Regional sediment management technical note ERDC/TN RSM‐18‐7. U.S. Army Corps of Engineers.
Sumi, T., Baoligao, B., & Morita, S. (2007). Characteristics of fine sediment discharge during sediment flushing of Unazuki Dam. Journal of
Hydroscience and Hydraulic Engineering, JSCE, 25, 99–106.
Sumi, T., Yoshimura, T., Asazaki, K., Kaku, M., Kashiwai, J., & Sato, T. (2015). Retrofitting and change in operation of cascade dams to facilitate
sediment sluicing in the Mimikawa River Basin. In Proceedings of 25th ICOLD Congress (pp. 597–611).
SWAN Team. (2021). Swan: Scientific and technical documentation. Version 41.31AB. Delft University of Technology. Retrieved from http://
www.swan.tudelft.nl
Tabios, III, G. Q. (2020). Reservoir sedimentation studies. In Water Resources Systems of the Philippines: Modeling studies (Vol. 4, pp. 165–209).
Springer. https://doi.org/10.1007/978‐3‐030‐25401‐8_6
TRMO. (2014). Flood early warning and flood control integration project under the jurisdiction of the Tenth River Management Office in 2014.
Tenth River Management Office, WRA, MOEA.
van Rijn, L. C. (2007). Unified view of sediment transport by currents and waves. I: Initiation of motion, bed roughness, and bed‐load transport.
Journal of Hydraulic Engineering, 133(6), 649–667. https://doi.org/10.1061/(asce)0733‐9429(2007)133:6(649)
van Rijn, L. C. (2013). Sedimentation of sand and mud in reservoirs in rivers. Retrieved from https://www.leovanrijn‐sediment.com/papers/
Reservoirsiltation2013.pdf
Wang, G., Wu, B., & Wang, Z. (2005). Sedimentation problems and management strategies of Sanmenxia Reservoir, Yellow River, China. Water
Resources Research, 41(9), W09417. https://doi.org/10.1029/2004WR003919
Wang, H.‐W., Kondolf, M., Tullos, D., & Kuo, W. C. (2018). Sediment management in Taiwan’s reservoirs and barriers to implementation.
Water, 10(8), 1034. https://doi.org/10.3390/w10081034
White, R. (2001). Evacuation of sediments from reservoirs. Thomas Telford.
Wohl, E. E., & Cenderelli, D. A. (2000). Sediment deposition and transport patterns following a reservoir sediment release. Water Resources
Research, 36(1), 319–333. https://doi.org/10.1029/1999wr900272
Xia, Q., & Liu, J. (2003). Sediment management at Naodehai reservoir. In Paper presented at World Water and Environmental Resources
Congress 2003. https://doi.org/10.1061/40685(2003)54
Yang, X. (2003). Manual on sediment management and measurement. In Operational hydrology report (Vol. 47). World Meteorological
Organization.
Yu, Y., Xuefa, S., Houjie, W., Chengkun, Y., Shenliang, C., Yanguang, L., et al. (2013). Effects of dams on water and sediment delivery to the sea
by the Huanghe (Yellow River): The special role of Water‐Sediment Modulation. Anthropocene, 3, 72–82. https://doi.org/10.1016/j.ancene.
2014.03.001
Zamora, J. (2018). Assessment of sediment handling strategies in the regulation pond of El Canadá hydropower plant, Guatemala. Master thesis.
Norwegian University of Science and Technology.
Zhang, W.‐Z., Shi, F., Hong, H., Shang, S., & Kirby, J. T. (2010). Tide‐surge interaction intensified by the Taiwan Strait. Journal of Geophysical
Research, 115(C6), C06012. https://doi.org/10.1029/2009jc005762
Zhou, Z. (2007). Reservoir sedimentation management in China. Lecture notes. International Research and Training Center on Erosion and
Sedimentation. Retrieved from http://www.irtces.org/nszx/zt/training2007/ppt/Lecture%202‐Zhou%20Zhide.pdf
Zhuang, S., Lu, X., Yu, B., Fan, X., & Yang, Y. (2021). Ascertaining the pollution, ecological risk and source of metal(loid)s in the upstream
sediment of Danjiang River, China. Ecological Indicators, 125, 107502. https://doi.org/10.1016/j.ecolind.2021.107502

HSUEH ET AL. 27 of 27