COASTAL AND ESTUARINE PLANNING FOR FLOOD AND EROSION PROTECTION USING INTEGRATED COASTAL MODEL

Yan Ding1, Keh-Chia Yeh2, Hung-Kwai Chen3, and Sam S. Y. Wang4

ABSTRACT This paper presents a practical engineering application of an integrated coastal model in coastal and estuarine planning for flood and erosion protection against hazardous hydrological forcing such as river floods, storm surges, waves, and high tides in an estuary in Taiwan. It demonstrates the advancement of the integrated modeling system to simulate hydrodynamic and morphodynamic processes in the estuarine area including rivers and adjacent coastal zones in responses to combined tides, waves, river flood, and winds. In order to identify the best engineering plan for the purpose of flood prevention and erosion protection, the integrated coastal models are applied to evaluate a number of engineering plans by simulating long-term variations of hydrodynamic and morphodynamic processes under a hypothetical storm-monsoon event which contains a 100-year storm and a three-month-long monsoon. Numerical results on flood water stages and morphological changes enable engineers to find the most desirable engineering plan for protecting the estuarine area from flood inundation and erosions. The integrated modeling system provides a comprehensive assessment tool for coastal and estuarine planning in complex hydrological and geomorphologic conditions. Keywords: Coastal and Estuarine Planning, Numerical Modeling, Coastal Flood, Erosion

INTRODUCTION Coastal floods during hazardous storms and hurricanes/typhoons can be devastating by causing flooding water inundations, severe coastal erosions, and casualties. Full understanding of their mechanism and accurate prediction of inundated water propagations and morphological changes is vital to flood management, erosion protection planning, and coastal environmental impact assessment. Integrated hydrodynamic and morphodynamic process modeling has become a necessary tool for planners and decision-makers to assess socio-economic and environmental impacts of hazardous hydrological forcing driven by tide, wave, river flood, wind, and sediment transport. Numerical simulations of coastal processes under various hydrological conditions can facilitate multiple-purpose engineering practices in developing cost-effective coastal flood management plans, as well as designing erosion control structures. In terms of process-integrated coastal models, simulation of coastal morphodynamic changes including shoreline evolutions, local scouring, and levee breaching has become feasible (e.g. Shimizu et al., 1996; Zyserman and Johnson, 2002, Tuan et al., 2008, Kuiry et al., 2010). In general, this was accomplished by sequentially computing wave field, flow field, and bed elevation changes. After a complete simulation cycle for hydrodynamic and morphodynamic processes, a new bathymetry will be fed back to affect the computations of the wave and current fields in the next time step. By this iterative procedure going through the wave-current-morphology models, the integrated modeling system is able to simulate the morphological process by using an empirical sediment transport model for the fine time-scale 1 Corresponding author: National Center for Computational Hydroscience and Engineering, The University of Mississippi, University, MS 38677, U.S.A. e-mail: ding@ncche.olemiss.edu 2 Department of Civil Engineering, National Chiao Tung University, Hsinchu, Taiwan. 3 Water Resources Planning Institute, Water Resources Agency, MOEA, Taichung, Taiwan. 4 National Center for Computational Hydroscience and Engineering, The University of Mississippi, University, MS 38677, U.S.A.

2

morphological process (e.g. Reniers et al., 2004; Ding et al., 2006b, Kubatko et al., 2006). At present, most existing storm-surge models can be used for simulating storm surge and coastal inundation, but some of them are incapable of simulating sediment transport and morphological change together with coastal hydrodynamic simulation, e.g. POM model (Blumberg and Meller, 1987), SLOSH model (Jelesnianski et al., 1992; SLOSH, 2008), SHORECIRC model (Svendsen et al., 2003ab). Recently, Pandoe and Edge (2008) applied the ADCIRC-2DTR model to study cohesive sediment transport in a ship channel in Matagorda Bay, Texas. Papanicolaou et al. (2008) presented a comprehensive review about sediment transport modeling and gave valuable suggestions on future model development for simulation of morphodynamic processes in rivers, coasts, and estuaries. Moreover, some researchers proposed several semi-empirical numerical approaches to assess long-term morphological changes on both the meso- and macro-scales in coastal inlets and tidal lagoons in estuaries, e.g., tide-averaging approach and Rapid Assessment of Morphology (RAM) (e.g. Roelvink, 2006). Apparently, these tidal-phase-averaging approaches can not capture the peak surge elevations and maximum flood inundation areas in storms/hurricanes; therefore, they aren’t appropriate for coastal flood protection planning. Complex and unsteady flow in river mouths and estuaries usually induce significant temporal/spatial morphological changes in a short storm period. To access impacts of extreme hydrological events such as hurricanes, storms, and high tides on coastal/estuarine areas with infrastructure, the most important capabilities of numerical models are to be able to accurately and robustly predict coastal flooding/inundation and coastal morphological changes due to these combined hydrological conditions over large-scale coastal/estuarine regions (Ding and Wang, 2008). The application to the simulations of large-scale and long-term morphodynamic processes due to extreme hydrological events in coasts and estuaries is a challenge to most existing numerical models. This paper presents an engineering application of CCHE2D-Coast, which is an integrated coastal and estuarine processes model (Ding et al., 2006b, Ding and Wang, 2008) to simulate hydrodynamic and morphodynamic responses to various hydrological conditions by typhoons, storm waves, river floods, and their combinations at an estuary located at the west coast of Taiwan (Figure 1). This estuary, called Touchien Estuary, has equally important coastal and estuarine processes driven by tides, waves, river inflows, and winds. The sediments can be transported into the estuary by means of river flows, tidal currents, wave breaking across the surf zone, and typhoon/storm surges. The coastal and estuarine morphodynamic processes are of multiple-scale motions and therefore very complex.

Figure 1 Interactions of various physical forcing in Touchien Estuary: the arrows indicate flow directions

driven by wave, tide, and river inflow.

Jiugang Is.

Ebb

Fengshan River

Touchien River

Flood

Wave

Tide

River Mouth Bar

Beiliao Is.

3

Due to hazardous weather conditions by typhoons and river floods, the local coastal community with a dense population is constantly threatened by coastal flooding and inundations, as well as channel refilling in harbors and property damage by shoreline erosions. Hence, several engineering conceptual plans for flood prevention and erosion protection in the area have been proposed by the local engineers. In order to identify the most desirable plan for the purpose of flood and erosion protection for the estuary, this engineering application project aims to assess the performance of these coastal/estuarine plans by using the integrated coastal model, CCHE2D-Coast, to simulate hydrodynamic and morphodynamic responses to the hypothetical extreme storm event and a long-term monsoon. For this project, before applying the CCHE2D-Coast model to evaluate the performance of the engineering plans, the reliability of the modeling results has been established through a comprehensive validation process in which historical morphological development scenarios are successfully simulated. It is confirmed that the numerical simulations reproduced complex but overall flow patterns and morphological changes in the estuary, e.g., bank overflow and coastal inundation, erosion/deposition, as well as breaching of river mouth sand bar. Then, the validated CCHE2D-Coast was utilized to simulate hydrodynamic and morphodynamic responses of the engineering plans to a hypothetical storm event containing a 100-year typhoon and a three-month monsoon. Numerical results about the highest water stages, flood flow propagation, and erosion patterns driven by the hypothetical extreme storm event, enable engineers to identify the most desirable plan for flood prevention and erosion protection in the local area. INTEGRATED COASTAL PROCESS MODEL – CCHE2D-COAST

The integrated coastal process model, CCHE2D-Coast, is applied to simulate the hydrodynamic and morphodynamic processes in responses to the given hydrological conditions with respect to storms/typhoons, waves, tides, and river floods so that the numerical results can be used for evaluating the proposed engineering plans for coastal and estuarine protections. This numerical software has been extensively verified and validated since it was developed in the National Center for Computational Hydroscience and Engineering at the University of Mississippi (e.g. Ding et al., 2006ab, Ding and Wang, 2008). It has been widely used for simulations of waves, currents, sediment transport, and morphological changes in different scales of coasts and estuaries for the purposes of coastal/estuarine erosion protection and flood water management. This software package contains three major submodels for simulating irregular wave deformations, tidal and wave-induced currents, sediment transport, and coastal morphological changes. As for the simulations of irregular waves, a multi-directional spectral wave transformation equation, with the diffraction effect terms, was adopted in the wave spectral module. The hydrodynamic module is capable of simulating tidal currents, river flows, and nearshore currents induced by short waves (i.e., wind-induced waves and swell waves). The morphodynamic module will compute morphological changes due to sediment transport under the conditions of the combined waves and currents. The module can take into account various coastal structures, e.g., groins, offshore breakwaters, artificial headlands, jetties, artificial reefs (submerged dikes in coasts), in computational domain. For the details of the model, one may refer to Ding et al. (2006b).

This integrated model for simulation of coastal estuarine morphodynamic processes has been built in a software package called CCHE2D (Jia et al., 2002), which is a general tool to analyze two-dimensional (2D) shallow water flows, sediment transport, and water quality, with natural flow boundary conditions. Similar to the CCHE2D hydrodynamic model, the three submodels in CCHE2D-Coast were discretized in a non-orthogonal grid system so that the models have more flexibility to simulate physical variables in complex coastal zones with

4

irregular coastlines. A time-marching algorithm proposed by Jia et al. (2002) was used to compute the tidal and wave-induced currents driven by typhoons and storm surges. A validated algorithm in CCHE2D for the treatment of wetting/drying area was directly used for predicting tidal flat variations and coastal inundations. For the verification and validation on CCHE2D-Coast for simulating waves, wave-induced currents, and morphological changes in coastal applications in various laboratory and field scales, one may refer to Ding et al. (2006ab).

MODEL VALIDATION IN TOUCHIEN ESTUARY Study Site and Computational Domain The study estuary called Touchien Estuary is located at the west coast of Taiwan Island. As shown in Figure 1, this estuary has a 1.0-km wide river mouth, a rivermouth bar, two islands inside the bay, and two rivers at upstream (Touchien and Fengshan Rivers). During storms or typhoons, this estuary has equally important hydrodynamic and morphological processes driven by tides, waves, and river floods. The sediments transport results from river flows, tidal currents, and wave breaking across the surf zone. The morphodynamic processes in the estuary are of multiple-scale motions and therefore complex. To investigate the hydrodynamic and morphodynamic responses to storms/typhoons and flood events in the estuary, a computational domain, as shown in Figure 2(a), was used in the simulation for model validation and engineering plan assessment. This non-orthogonal structural grid was generated to cover the entire estuarine and coastal areas. Figure 2(b) shows a close-up of the non-orthogonal mesh in the estuarine area in which several interesting sites are marked as monitoring stations for numerical result outputs. Through simulating the morphological changes from 2004 to 2006, the CCHE2D coastal model was validated. To do so, the bathymetrical data for the estuary obtained from the observations in 2004 were used to generate bed elevations in the computational mesh.

0 1km 2km

BED (m)

543210

-1-2-3-4-5-10-15-20

N

Touchien

Fengshan

Offshore

Rivermouth Bar Head

Mid of Island

0 1k m

Figure 2 A non-orthogonal mesh covering the estuary and coast

Hydrological and Morphological Conditions The boundary conditions for modeling the morphological processes include the hydrographs of inflows at the upstream of the two rivers, tidal elevations at the offshore boundary, incident wave properties from offshore, wind forcings, and sediment properties in the estuarine/coastal area and the sediment transport rates from the two rivers. The offshore boundary conditions, tides, waves, winds, etc., were provided by a research group at the

(a) (b)

5

National Cheng Kung University, Taiwan. The conditions includes all the 9 typhoons that occurred during the period from 2004 to 2006 (Table 1). The hydrographs and the sediment properties for the two river inflows during the same period of the computation were provided by another research group in the National Chiao Tung University, Taiwan. Two hydrographs at the two inlet cross-sections of the Fengshan and Touchien rivers have been obtained by performing a one-dimensional (1D) river simulation from far upstream of the two rivers to the estuary under the same hydrological condition of the flood waters during the computational duration from 2004 to 2006. The peak discharge was 5,343 m3/s in the Touchien river in Typhoon Matsa at 14:00, 8/5/2005.

Table 1 Nine typhoon events occurred from 2004 to 2006 for the present study

Typhoon Name Starting Date End Date Mindulle June 26, 2004 July 10, 2004

Aere Aug. 18, 2004 Sept. 2, 2004 Haima Sept. 5, 2004 Sept. 19, 2004 Haitang July 12, 2005 July 26, 2005 Matsa July 29, 2005 Aug. 12, 2005 Talim Aug. 27, 2005 Sept. 2, 2005

LongWang Sept. 23, 2005 Oct. 7, 2005 Bilis July 3, 2006 July 17, 2006

Kaemi July 19, 2006 Aug. 1, 2006

The boundary conditions of tidal elevations and wave properties at the offshore in the three-year simulation were given by using a regional storm-surge model (a POM model) and the SWAN wave model (SWAN 2007). The wave properties, i.e., significant wave heights, peak periods, and mean directions, were provided by the measurements and the extracted results from the simulations by the SWAN model. The significant wave heights were basically lower than 2.0 m in most typhoons and flood events at the coast. However, the maximum wave height offshore reached 4.21 m during Typhoon Bilis in July, 2006. The wind speed data during the simulations were also provided by the regional storm-surge model.

According to the grain size measurements, a uniform grain size, d50 = 0.2 mm, was used for representing the coastal sediments in the domain. A total load sediment transport formulation was used to calculate sediment fluxes and morphodynamic changes (Ding et al. 2007). The bottom roughness coefficient, i.e., Manning’s n, was set to 0.025 on the sea bed, and 0.033 on the river bed based on the 1D river simulation model.

To simulate the random waves, the Bretschneider-Mitsuyasu (B-M) spectrum (Ding et al., 2006b) was used to specify the spectral wave inputs from offshore. The lower and upper frequency bounds were set to 0.05 Hz and 10 Hz, respectively. The frequency interval 0.4975 Hz (i.e. 21 frequency bins) and the angle interval 5.0O (i.e., 37 directional bins between –90O and +90O) were adopted. The effects of wave breaking and wave diffraction (Ding et al. 2006b) were considered in the irregular wave simulations. The time interval for the hydrodynamic and morphodynamic modeling was 2 s. The wave field was updated every hour hydrodynamic and morphological computation. Model Validation Results

The initial conditions of the currents and water elevations were obtained from the steady flow with the constant discharges and the first wave action under a non-tide mean sea level (MSL) situation. The computed flow fields include the interaction of tidal currents, nearshore currents, and river floods from upstream. Therefore, the currents represent the highly

6

complicated multi-scale hydrodynamics in the estuary, which contains coastal flooding, river mouth bar breaching, and interactions of tidal and nearshore currents. The computed water elevations indicate that the high flood waters in Typhoons Aere, Matsa, and Talim could inundate the north bank of the Touchien river, and overflow the Beiliao Island (an uninhabited island shown in Figure 1). The long-term simulation results about the water elevations reproduced the coastal flooding during the three-year period in which the nine typhoon events occurred. The highest water stage, 12.27 m, occurred at the Touchien river inlet cross-section at the peak discharge of Typhoon Matsa. At the same time, the water stage at the Beiliao Island was 5.78-m high, and then the island was submerged in the flood waters.

The coastal morphological model is validated by comparing the computed morphological changes with the measured bed elevation changes. The measurement area in 2006 covered the Touchien estuarine area and the tidal reaches of the two rivers. The measured bed elevation changes were obtained by comparing their measurements in 2006 with those in 2003. To validate the coastal morphodynamic process model, calibrations of empirical parameters related to the morphodynamic model were conducted. To do so, the computed bed elevation changes at ten selected stations as shown in Figure 3(a) were compared with the measured bed changes at the same locations. Meanwhile, only two empirical parameters were calibrated: an empirical parameter of sediment transport rate (Bw) in the Watanabe’s total load formulation, and the coefficient of downslope gravitational effect (ε) in the morphological change equation (see the definition in Ding et al., 2006b). By testing the morphodynamic simulations for more than 12 runs, the site-specific calibrated parameters, Bw = 3.0, and ε = 10.0, were obtained. The comparisons of the bed changes at the ten stations computed by using the calibrated parameters are shown in Figure 3 (b); it indicates that an excellent agreement between the measurements and observations was obtained at these selected stations in the estuary.

To compare the spatial distribution of the bed changes in the entire estuary, the measured morphological changes over the three years are shown in Figure 4 (a), in which the red color indicates deposition of sand, and the blue means erosion of bed. The simulated morphological changes through the three-year typhoon events are plotted in Figure 4 (b) with the same legend in Figure 4 (a). By visually comparing the two figure measurements and simulations, the following common features on deposition and erosion are found:

(1) The long-term morphodynamic simulations correctly reproduced the erosion and breaching that occurred in the river mouth bar;

(2) The simulations produced a similar erosion pattern at the head of the Beiliao island, even though the computed erosions are underestimated; (3) The simulated sand depositions in the south bank of the river mouth show a consistent deposition pattern with the observations; (4) The morphodynamic simulations produced a similar offshore bar development; yet, the size of the simulated offshore bar is larger than the observations; and (5) The simulated morphological changes in the Touchien River show almost the same size and locations of deposition and erosion, e.g., the erosion at the left bank and the deposition at the right bank near the Jiugang Island. However, the simulations in the Fengshan River reach show overestimated deposition. Finally, the accuracy of the model to simulate the morphological changes is investigated by comparing the bed changes in all the measurement stations in the estuary. Ding et al. (2008) reported the measured and simulated bed changes at all nodes for which the bed changes are not less than 5 cm. Through the comparisons in all the nodes, the average absolute error of bed changes was found to be ±37.0 cm. Even though comparisons are made in this large estuarine area including the river reaches, the predication accuracy is still in a reasonable range which is acceptable for the engineering practical applications.

7

dz

43.63.22.82.421.61.20.80.40

-0.4-0.8-1.2-1.6-2-2.4-2.8-3.2-3.6-4

N

0 1 2kmTouchien

Fengshan

Computed Bed Elelvation Changes

St. 2

St. 1

St. 3

St. 4

St. 5

St. 6

St. 7 St. 8

St. 9

St. 10

-8

-6

-4

-2

0

2

4

6

8

-8 -6 -4 -2 0 2 4 6 8

Computed Bed Change (m)

Mea

sure

d Be

d C

hagn

e (m

)

(a) Selected monitoring stations (b) Comparison of bed changes at the stations

Figure 3 Comparison of bed changes at selected stations

(a) Measured bed changes (b) Computed bed changes

Figure 4 Comparison of bed changes between measurements and simulations

MODEL APPLICATION TO COASTAL AND ESTUARINE PLANNING There are six engineering conceptual plans proposed to regulate the Touchien Estuary for the purpose of flood prevention and erosion protection. Including the case of the present estuary (status quo, Case 1), a total of seven proposed cases have to be evaluated by simulating the flows and morphological changes over a long-term period. The detailed descriptions of the engineering plans are shown in Figure 5. The engineering countermeasures adopted against flooding and erosion are (1) installation of a 7.0-m high dike to protect the south bank of the Touchien River in Cases 2, 3, 4, and 7; (2) dredge of the channel of a branch in the north of Jiugang Island in Cases 3, 4, and 6; (3) removal of Beiliao Island in the estuary in Cases 6 and 7; and (4) installation of a jetty to separate waters in the river mouth in Case 5. In order to investigate the performance of these engineering plans, simulations of hydrodynamic and morphodynamic responses to a hypothetical extreme typhoon event and a monsoon event were performed.

As shown in Figure 6, the hypothetical computational period for evaluation of the

-10

-8

-6 -4

-4-4

-2

-2

-2

-2

-2

-2

0

0

00

00

0

0

0

00

2

2

2

2 2

2

4

4

DZ (m)

21.81.61.41.210.80.60.40.20.050

-0.05-0.2-0.4-0.6-0.8-1-1.2-1.4-1.6-1.8-2

N

0 1 2km

Touchien

Fengshan

Measured Bed Elelvation Changes

Beiliao Is.

River Mouth Bar

South Bank Jiugang Is.

-10

-8

-6 -4

-4-4

-2

-2

-2

-2

-2

-2

0

0

00

0

0

0

0

0

0

0

2

2

2

2 2

2

4

4

DZ (m)

21.81.61.41.210.80.60.40.20.050

-0.05-0.2-0.4-0.6-0.8-1-1.2-1.4-1.6-1.8-2

N

0 1 2km

Touchien

Fengshan

Computed Bed Elelvation Changes

Initial Bed Elevation

Erosion

8

plans contains a typhoon and a monsoon. This extreme typhoon event with a 100-year return period was determined beforehand by using the historical data of typhoons passing through the island from 1897 to 2001. Driven by the 100-year typhoon, a hypothetical 100-year flood was assumed to take place in the two rivers upstream for 56 hours. Two hydrographs representing the 100-year flood at the two inlet cross sections of the two rivers were obtained by a one-dimensional river flood simulation covering two river reaches from a faraway mountain area to the estuary section. The maximum peak discharges in the Touchien River reached 7,072 m3/s, in 13 hours, and another peak flood in the Fengshan River at 1,872m3/s, flowed down to the estuary one hour later. In the monsoon period, the two small discharges for a dry season were assumed to apply to the two rivers, i.e., 4.75 m3/s in the Touchien River and 2.22 m3/s in the Fengshan River.

Figure 5 Engineering Plans for Flood and Erosion Protection in Touchien Estuary

0

2000

4000

6000

8000

0 10 20 30Hour

Dis

char

ge (C

MS)

Touchien RiverFengshan River

100-year flood

0

2

4

6

8

10

30 50 70 90Hour

Dis

char

ge (C

MS)

Touchien RiverFengshan River

Monsoon

Figure 6 Hypothetical hydrographs to represent a 100 year flood and a dry season in a monsoon

The tidal elevations, during Typhoon Haitang (7/21-7/23/2005) and a three-month

monsoon from 12/01/2006 to 3/1/2006, were used as the offshore tidal boundary conditions

Case 2 Case3 Case 4

Case 5 Case 6

• Dike (+7.0-m high) • Remove Beilaio Island • Dredge of Channel • Land Reclamation

• Jetty to separate the two rivers

• Dike (+7.0-m high) • Dredge of Channel

• Dike (+7.0-m high) • Dredge of Channel • Land Reclamation

• Dike (+7.0-m high) • Land Reclamation

• Dike (+7.0-m high)

• Remove Beiliao Island

­W­d¸}³ö Dike

Dredge

Dike

苦苓腳堤

苦苓腳堤防

右股河道疏浚

Case 7

Dredge

Dike

Jetty Dike

Dredge

Remove Island

Dike

Remove Island

9

in the hypothetical typhoon-monsoon event. The largest tidal range of the spring-neap tides was almost 5 m. The wave properties, i.e., significant wave height, period, and wave mean direction, for the offshore boundary conditions of the CCHE2D-Coast model were extracted from the computed results by SWAN wave model in a large regional grid. The maximum storm wave heights in the events reached to almost 7.0 m. The wind effects were also included in the hydrodynamic simulations by using the measured data of wind speeds and directions. Figure 7 (a) and (b) show respectively a computed wave field (significant wave heights and mean directions) and currents at the time of the peak flood (t=12 h). Figure 7 (a) shows a computed wave field by the wave spectral model in the estuary including the installation of structures (breakwaters and dikes). Figure 7 (b) presents the computed currents and water elevations driven by the wave field shown in Figure 7 (a), tides, river floods, in which the breaching and the overbank flooding were predicted at the rivermouth bar. The currents at the peak flood indicate a combined flow field at the moment by river flood, longshore current, and tidal current. Figure 8(a) shows another flow field and the corresponding estuary shape at an ebb tide after this 100-year flood and storm has gone. It is found that the rivermouth bar breaching and rivermouth widening occurred due to the extreme flood. Figure 8 (b) depicts the computed morphological changes in the estuary after the 100-year storm and flood, which quantitatively shows the coastal erosions in the river mouth, the breaching in the river mouth bar, the river mouth widening, and the offshore bar development due to the river sediment flushed down to the offshore.

Figure 9 further depicts the time histories of water elevations at the seven monitoring stations, of which locations are shown in Figure 9(a). Since the peak flood was coincident with a high tide at the coast, it induced severe inundations in the estuary as shown in Figure 7(b). The impacts of the hydrological forcings can be found from the variations of the water elevations: For example, by comparing the water elevations at the peak flood (t=12h), it is found that the storm resulted in an approximately 1.5-m high surge at the river mouth. However, the water stage at the Touchien upstream can reach to 14 m due to the upstream flood, storm surge, wave set-up, and tide.

The highest water stage is an important index to evaluate the impact of storm/flood waters in the estuary. By selecting several representative locations in the estuary, the highest water stages were extracted from the time histories of the computed water elevations. In Case 1 (the status quo case), Figure 10 presents the highest (peak) water stages at Nanliao, where is the most vulnerable location for flooding. The red dash line in the figure is the design dike high (+7.0m) at the south bank of the Touchien River.

Based on the evolutions of the bed elevations at the selected stations in the estuary, the accumulated seasonal bed changes after the two events, i.e., 100-year storm and 3-month monsoon, are calculated and shown in Figure 11. The features of the bed changes in the estuary are summarized as follows:

(1) The storm can cause river mouth erosion, and deposition in the monsoon; (2) The periodical morphological changes occur in the river mouth. This predicted

morphodynamic mechanism in the river mouth is in virtual agreement with the coastal and estuarine morphological processes in the area;

(3) After the whole period, the net erosion is predicted in Beiliao, Nanliao, and two river flow inlets; and

(4) The net deposition occurs in Jiugang channel, created solely by the storm. It indicates that the Jiugang channel has a potential risk of refilling by the sediment conveyed by the flood water from the upstream.

10

Hs

43.63.22.82.421.61.20.80.40

N

0 1 2kmTouchien

Fengshan

Computed Wave Heights and Directions

Hs = 2.0m

3

4

4

5

6

8

10

WL (m)

121110987654321

N

0 1 2kmTouchien

Fengshan

Current = 1.0 m/s

(a) Computed wave heights and mean directions (b) Computed currents and water elevations

Figure 7 Computed currents and water elevations at the flood peak (t=12h)

Vel (m/s)

21.81.61.41.210.80.60.40.20

N

0 1 2kmTouchien

Fengshan

Current = 1.0 m/s

-30

-25

-20

-20

-15

-15

-10

-5

-5

-5

-4

-4

-4

-3

-3

-2

-2

-1

-1

0

0

0

1

1

1

1

1

11

1

2

2

2

2

3

3

3

3

DZ

0.50.40.30.20.10

-0.1-0.2-0.3-0.4-0.5

N

0 1 2km Touchien

Fengshan

Computed Sediment Fluxes, Bed Elelvations, and Changes

Sediment Flux = 0.001 m2/s

(a) Computed ebb current (b) Bed changes, bed elevations, sediment fluxes

Figure 8 Computed currents, bed changes, bed elevations, and sediment fluxes at the end of the 100-year flood and storm (t=27 h)

-20

-10

-15

-5

-4 -3 -2

-1

0

2

4

3

01

2

0

1

0

BED

543210

-1-2-3-4-5-10-15-20

N

Touchien

Fengshan

Offshore

Rivermouth Bar

Nanliao

0 1000 m

Jiugang Channel

Beiliao Island

Hours

Wat

erE

leva

tions

(m)

0 10 20 30 40 50-5

0

5

10

15

20

Touchien InletOffshoreRiver MouthBeiliao IslandFengshan InletNanliaoJiugang Channel

(a) Monitoring Stations (b) Time Histories of Water Elevations

Figure 9 Time histories of water elevations at seven monitoring stations

Longshore current

River flood current

Tidal current

Beiliao Is.

11

14.37

3.33 3.83

6.18

8.196.67

10.49

0

2

4

6

8

10

12

14

16

Touchien Inlet Offshore River Mouth Beiliao Island Fengshan Inlet Nanliao JiugangChannel

Wat

er S

tage

(m)

Design Dike High (+7.0m) at Nanliao

Figure 10 Highest water stages in Case 1 at seven monitoring stations in the 100-year storm

Figure 11 Seasonal bed changes in Case 1

All six engineering plans have been investigated by simulating the hydrodynamic and morphodynamic responses to the above-mentioned hypothetical storm-monsoon event. In order to find out the best plan for flood prevention purposes, the highest water stages at the selected monitoring stations for all the cases are shown in Figure 12. Some preliminary remarks are given as follows: Removal of Beiliao Island can lower the flood water elevations. This engineering plan has better flood protection than Case 4 (the Jiugang Channel dredged). The highest water stages in the monitoring stations in Case 7, i.e., Beiliao, Nanliao, Jiugang, and Touchien upstream, are lower than these in Case 4. They are also the lowest in all the cases. In comparison with Case 4, it is found that the removal of Beiliao Island can make the highest water stage lower by 0.63m at Beiliao, 0.79m at Nanliao, 0.24m at Jiugang, and 0.19m at Touchien upstream. In Case 4, dredging the Jiugang channel and diverting flood waters can slightly alleviate the flooding downstream (0.13m lower at Nanliao; 0.10 m lower at Beiliao). By comparing this effect with the dredged Jiugang channel in Case 4, the removal of the Beiliao Island can effectively reduce water stages in the estuary. Therefore, Case 7 (Dike + island removal) could be the best plan over the seven cases (including Case 1 which is doing nothing) to prevent potential flooding of the Touchien Estuary. The morphological changes in the computational period are divided into three parts: (1) the net bed change over the entire period induced by the storm and monsoon; (2) the bed change induced by the 100-year storm only; and (3) the bed change caused by the monsoon only. As shown in Figure 13, both two seasons (storm and monsoon) could induce bed erosions on the upstream Touchien and Fengshan inlet cross-sections in almost all the cases. Erosion by the storm flood is especially prominent in the morphological changes upstream. It means that the upstream river beds are degraded in all the periods. The bed elevation changes in the station of Jiugang Channel in Figure 13 indicate that the north channel of the Jiugang Island is always aggraded in both storm and monsoon seasons. However, in Cases 3, 4, and 6, in which the Jiugang channel is planned to be dredged, the

12

storm flood water can trigger rapid deposition (more than 2 m) on the dredged channel bed. With the exception of the cases that the land reclaimed by placing a 7-m high dike in Nanliao (hard structure), the river bed at Nanliao is slightly eroded in Cases 1, 4, 5, and 7 in the monsoon; the bed degradation is then slow, approximately 15 cm over the whole season. However, in Case 7, in which the Beiliao Island is planned to be removed, the flood in the storm can make more deposition on the dredged river bed near Nanliao than the monsoon does. Similarly, because of the removal of Beiliao Island, only Case 6 and Case 7 can cause deposition on the river bed near Beiliao where the bed is dredged. All the other cases have erosions on the bed at the same location. The complicated morphological changes occurring in the river mouth indicate that the bed changes on the river mouth bar are periodical (see Figure 1). In Case 1 – Case 4, the storm flood water always leads to erosions on the river mouth, but the tidal currents and longshore currents in the monsoon make depositions. The net bed aggradation in the station can be found in Case 1 and Case 2. Case 3 resulted in net bed degradation. Only Case 4 shows almost no bed changes. In contrast, in Case 5, the storm creates deposition on the river mouth, but the monsoon causes erosion. Due to the jetty, Case 5 shows an opposite trend of the bed change on the river mouth. In Case 6 and Case 7, due to the removal of the Beiliao Island, two seasons result in erosions on the river mouth bar.

Figure 12 Comparisons of highest water stages in the seven cases at six selected stations

By summarizing the above analysis of flood water stages and bed changes for all the seven cases, it is found that the planned dike with the Beiliao Island removed in Case 7 can better prevent inundation of the estuarine area of the Touchien River. As for morphological changes, it is found that the bed changes upstream of the rivers in Case 7 have a similar pattern to Case 4; whereas, the bed changes inside the estuary are similar to the changes in Case 6. However, the simulation results show that the storm flood brings sand down to the estuary, and sands are readily deposited into the dredged river bed due to the removal of Beiliao Island. Therefore, the longer simulation in Case 7 may be

-20

-10

-15

-5

-4 -3 -2

-1

0

2

4

3

01

2

0

1

0

BED

543210

-1-2-3-4-5-10-15-20

N

Touchien

Fengshan

Offshore

Rivermouth Bar

Nanliao

0 1000 m

Jiugang Channel

Beiliao Island

Touchien Inlet

14.37

14.61 14.6714.51

14.4214.51

14.32

14.00

14.20

14.40

14.60

14.80

1 2 3 4 5 6 7

Case No.

Wat

er E

leva

tion

(m)

Fengshan Inlet

8.19 8.21 8.21 8.19 8.19 8.18 8.19

8.00

8.10

8.20

8.30

1 2 3 4 5 6 7

Case No.

Wat

er E

leva

tion

(m)

Jiugang Channel

10.49

11.07 11.1410.81

10.5810.80

10.34

9.50

10.00

10.50

11.00

11.50

1 2 3 4 5 6 7

Case No.

Wat

er E

leva

tion

(m)

Nanliao

6.67

8.13 8.06

6.62 6.89 7.10

5.83

5.0

6.0

7.0

8.0

9.0

1 2 3 4 5 6 7

Case No.

Wat

er E

leva

tion

(m)

Beiliao Island

6.186.97 6.87

6.106.42

5.53 5.47

5.00

6.00

7.00

8.00

1 2 3 4 5 6 7

Case No.

Wat

er E

leva

tion

(m)

River Mouth

3.83 3.85 3.843.77 3.74

3.82 3.78

3.503.603.703.803.904.00

1 2 3 4 5 6 7

Case No.

Wat

er E

leva

tion

(m)

13

needed in the future since this hypothetical storm-monsoon long-term event may not be long enough to the dynamic equilibrium morphology in the estuary.

Figure 13 Seasonal bed changes of the seven cases at selected stations

CONCLUSIONS

This paper presents a practical engineering application of an integrated coastal model, CCHE2D-Coast, in coastal and estuarine planning for flood and erosion protection against hazardous hydrological forcing such as river floods, storm surges, waves, and high tides in Touchien Estuary in Taiwan. It demonstrates the advancement of the integrated modeling system to simulate coupled hydrodynamic and morphodynamic processes in a large area of the estuary including rivers and adjacent coastal zones, which are driven by combined tides, waves, river flood, and winds.

To find the best coastal/estuarine planning for flood and erosion protection, based on the previous simulation study on the coastal flood and morphology in the status quo estuary, the local engineers collaborating with numerical modelers have developed a totally six coastal/estuarine design/planning cases, which were based on conventional engineering means, including hard structure installations (e.g., dikes, jetties, etc) and soft engineering approaches (e.g., channel dredging, removal of island in the bay).

Before applying the model to evaluate coastal/estuarine design plans, model validation has been done by simulating hydrodynamic and morphodynamic processes over a series of historical typhoon events happened in between 2004 and 2006. Numerical simulations reproduced complex but overall flow patterns and morphological changes in the estuary, e.g., bank overflow and coastal inundation, erosion/deposition, as well as breaching and scouring of the river mouth sand bar.

Touchien Inlet

-1

-0.5

0

0.5

1 2 3 4 5 6 7

Case No.

Bed

Cha

nges

(m)

By Storm and Monsoon By Storm By Monsoon

Fengshan Inlet

-0.25

-0.2

-0.15

-0.1

-0.05

01 2 3 4 5 6 7

Case No.

Bed

Ele

vatio

n C

hang

es

(m)

By Storm and Monsoon By Storm By Monsoon

Jiugang Channel

00.5

11.5

22.5

1 2 3 4 5 6 7

Case No.

Bed

Cha

nges

(m)

By Storm and Monsoon By Storm By Monsoon

Nanliao

-0.2

-0.1

0

0.1

0.2

0.3

1 2 3 4 5 6 7

Case No.

Bed

Cha

nges

(m)

By Storm and Monsoon By Storm By Monsoon

Beiliao Island

-2

-1.5

-1

-0.5

0

0.5

1 2 3 4 5 6 7

Case No.

Bed

Cha

nges

(m)

By Storm and Monsoon By Storm By Monsoon

River Mouth

-1

-0.5

0

0.5

1

1 2 3 4 5 6 7

Case No.

Bed

Ch

ange

s (m

)

By Storm and Monsoon By Storm By Monsoon

14

Then, the validated models were applied to simulate hydrodynamic and morphodynamic responses in six engineering conceptual plans to a hypothetical long-term hydrological forcing, a storm-monsoon event. This event, containing a 100-year storm (representing extreme weather) and a three-month-long monsoon (representing fair weather), provides a hypothetical condition to study the impact of short-term extreme event due to storm surge and river floods, as well as the response of morphological changes to a relative long (three months) monsoon. By the systematical simulations of the seven cases (including the status quo case), the maximum (highest) flood water stages corresponding to the storm and the long-term bed changes due to the long-term monsoon fair weather were obtained. The highest flood stages are used to confirm the capability of the flood protection of the engineering plans; the long-term morphological changes are for evaluation of coastal erosion (including deposition) in the estuarine area. The intercomparisons of the maximum water stages and bed changes revealed that Case 7, i.e. using dikes and removing Beiliao Island, could be the most desirable engineering plan or the best choice in the proposed cases.

Hence, in terms of the integration numerical modeling and practical engineering knowledge on hydrological, morphological, meteorological, and geomorphic conditions, this study demonstrates a general procedure to achieve the best coastal/estuarine planning against flood and erosion. The planning assessment procedure may briefly conclude as follows: (1) study the historical data about hydrology and morphology of the engineering site, (2) select the study area and computational domain, and generate numerical mesh to represent coastal/estuarine geometry and infrastructure (dikes, jetties, harbors, etc.), (3) validate the integrated numerical models including hydrodynamic (wave and current) and morphodynamic (sediment transport and morphological changes) models, (4) propose coastal/estuarine design plans using the engineering approaches including hard structures and soft engineering means, (4)determine a hypothetical hydrological condition (or event) based on the historical data to represent extreme event (storm, typhoon/hurricane, river flood), (5) simulate the hydrodynamic and morphodynamic impacts in the proposed design plans, and (6) intercompare the numerical results on flows (water stages and currents) and morphological changes, and finally identify the best choice for the flood and erosion protection plan. ACKNOWLEDGMENTS

This work was a product of the collaborative research with National Chiao Tung University, and was sponsored by Water Resources Planning Institute, Water Resources Agency, MOEA, Taiwan. Especial appreciation is expressed to Mr. Chin-Yen Tsai in the National Cheng Kung University in Taiwan for providing the data for boundary conditions. REFERENCES Blumberg, A. F. & Mellor, G.L. (1987), “A description of a three coastal ocean circulation

model,” In: Three Dimensional Coastal Ocean models, Vol.4, Ed. By N. Heaps, American Geophysical Union, Washington, D.C.

Ding, Y., Jia, Y., and Wang, S. S.Y. (2006a), “Numerical modeling of morphological processes around coastal structures,” Proc. of ASCE-EWRI Congress 2006, Omaha, NE, May 21-25, 2006.

Ding, Y., Wang, S. S. Y., and Jia, Y. (2006b), “Development and validation of a quasi three-dimensional coastal are morphological model,” Journal of Waterway, Port, Coastal, & Ocean Engineering, ASCE, 132(6), 462-476.

Ding, Y., Yeh, K.-C., Chen, H.-K., and Wang, S.S.Y. (2007), “Simulations of morphodynamic changes due to waves and tides in an estuary using CCHE2D-coast model,” Proc. of ASCE-EWRI Congress 2007, May 15-19, 2007, Tampa, Florida.

15

Ding, Y., and Wang, S. S. Y. (2008), “Development and application of coastal and estuarine morphological process modeling system,” Journal of Coastal Research, Special Issue No. 52, pp127-140.

Ding, Y., Yeh, K.-C., Chen, H.-K., and Wang, S. S. Y. (2008), “Validation of a coastal and estuarine model for long-term morphodynamic simulations driven by tides, storms, and river floods,” In: World Environmental and Water Resources Congress 2008, Ed. by R. W. Babcock, Jr., and R. Walton, ASCE, May 12-16, 2008, Honolulu, Hawaii

Jia, Y., Wang, S. S. Y., and Xu, Y. C. (2002), “Validation and application of a 2D model to channel with complex geometry,” Int. J. Computational Engineering Science, 3(1), 57-71.

Jelesnianski, C., Chen, J. and Shaffer, W. (1992) ‘SLOSH: sea, lake, and overland surges from hurricanes’, NOAA Technical Report NWS 48, Silver Spring, MD.

Kubatko, E.J., Westerink, J.J., and Dawson, C. (2006), “An unstructured grid morphodynamic model with a discontinuous Galerkin method for bed evolution,” Ocean Modelling, 15, 71–89.

Kuiry, S. N., Ding, Y., and Wang, S. S. Y. (2010), “Modeling coastal barrier breaching flows with well-balanced shock-capturing technique,” Computers & Fluids, 39(10), pp2051-2068.

Reniers, A.J.H.M., Roelvink, J.A., and Thornton, E.B. (2004), “Morphodynamic modeling of an embayed beach under wave group forcing,” J. Geophys. Res., 109, C01030, doi:10.1029/2002JC001586.

Pandoe W.W. and Edge, B. L. (2008), “Case study for a cohesive sediment transport model for Matagorda Bay, Texas, with coupled ADCIRC 2D-Transport and SWAN Wave Models,” Journal of Hydraulic Engineering, ASCE, 134(3), 303-314.

Papanicolaou, A. N., Elhakeem, M., Krallis, G., and Edinger, J. (2008), “Sediment Transport Modeling Review – Current and Future Developments,” Journal of Hydraulic Engineering, ASCE, 134(1), 1-14.

Roelvink, J.A. (2006), “Coastal morphodynamic evolution techniques,” Coastal Engineering, 53, 277-287.

Svendsen, I. A., Haas, K., and Zhao, Q. (2003a), “Quasi-3D nearshore circulation model SHORECIRC – Version 2.0,” Technical Report, Center for Applied Coastal Research, University of Delaware, Newark, DE19716.

Svendsen, I. A., Qin, W., and Ebersole, B.A. (2003b), “Modelling waves and currents at the LSTF and other laboratory facilities,” Coastal Engineering, 50(1-2), 19-45.

Shimizu, T., Kumagai T., and Watanabe A. (1996), “Improved 3-D Beach Evolution Model Coupled with the Shoreline Model (3D-SHORE),” Proc. 25th Conf. On Coastal Eng.,ASCE, 3, pp.2843-2856.

SLOSH (2008), http://www.nhc.noaa.gov/HAW2/english/surge/slosh.shtml, <Jan. 16, 2008> Smith J. and Cialone M. (2000), “Waves and Currents: STWAVE,” Technology-Transfer

Workshop, Feb. 1–2, 2000, Hilton Airport Melbourne, Florida (http://cirp.wes.army.mil/cirp/presentations/ws2000/stwave/).

SWAN (2007), http://vlm089.citg.tudelft.nl/swan/index.htm, Delft University of Technology, Jan. 10, 2007.

Tuan, T. Q., Stive, M.J.F., Vergagen, H.J., and Visser, P.J. (2008), “Process-based modeling of the overflow-induced growth of erosional channels,” Coastal Engineering, 55, 468–483.

Zyserman, J.A., and Johnson, H.K. (2002), “Modelling morphological processes in the vicinity of shore-parallel breakwaters,” Coastal Engineering, 45(3), 261-284.