Coastal Erosion Dynamics and Protective Measures in the Vietnamese Mekong Delta

Abstract

The dynamic shifts in shorelines due to erosion and deposition have become a significant challenge in coastal zone management, particularly in the context of climate change and rising sea levels. This paper evaluates the shoreline protection and efficiency of various wave-reducing breakwaters in the Vietnamese Mekong Delta. The delta exemplifies the coastal erosion issue faced by deltas worldwide. Landsat satellite images were used to establish a coastal development map for the period 2000 to 2022. The wave data in front and behind the breakwaters were analyzed to assess the wave reduction efficiency of various breakwater structures. Our results reveal that coastal erosion is deeply concerning, with almost 40% of the coastline experiencing severe erosion. Hotspot areas have been observed to reach annual erosion rates of nearly 95 m per year. The majority of provinces have adopted protective measures, with 68% of affected shorelines protected to some degree. Our results show breakwaters to be highly effective in reducing wave height, with a 62% reduction in waves reaching the shore. The process of creating offset has taken place in the area from the breakwater back to the mainland, with the rate of increase in compensation also quite fast at up to 3.1 cm/month. The stability of the pile–rock is very high; however, it is necessary to add rock to compensate for the settlement of the rock part.

1. Introduction

The Vietnamese Mekong Delta (VMD) is the twelfth-largest delta in the world and a key economic region of Vietnam, including 19% of the population, contributing around 15% to the national GDP, and accounting for over 50% of rice production, 65% of aquaculture, and 70% of fruit production in the country [1,2]. Generally, the VMD experiences two distinct seasons—the dry season, with a prevailing northeast wind, and the rainy season, with a southeast wind—under the influence of the Asian monsoon climatic regime [3]. Despite its relatively recent formation, the VMD supports approximately 17.3 million inhabitants and plays a vital role in global food security [4,5]. Identified as one of the world’s most vulnerable deltas in terms of climate change, the region is heavily impacted by saltwater intrusion, droughts, floods, erosion, and subsidence. These challenges threaten the livelihoods of over 17 million people and affect the development of one of Vietnam’s most important and strategic regions [6,7,8]. While The VMD exemplifies the issue of coastal erosion in many tropical deltas, it is also facing challenges related to erosion risks and alterations in flow regimes, attributable to anthropogenic and climate-related influences [9,10]. Sediment fluxes reaching the ocean and the influence of waves and currents are important in redistributing the river’s sediment supply to the southwest [11].

The expansion of the delta towards the sea and the formation of the Ca Mau Peninsula (Figure 1) to the southwest were caused by extensive river sediment fluxes over a period of 3500 years [12]. Over the course of 6000 years, the VMD has expanded at an average rate of 7 km²/year due to sedimentation from the Mekong River, resulting in the formation of the Ca Mau Peninsula and the southwest region over a span of 3500 years [13]. However, recent investigations [14] have revealed a shift in sediment transport and deposition within the VMD since 2005, marked by a gradual decrease in shoreline growth of only 1.4 m from 2005 to 2015. Additionally, studies have highlighted instances of land subsidence and channel erosion in the VMD [15]. Human activities have significantly impacted the VMD since the mid-20th century. These activities include reduced river fluxes due to damming and sand mining, land subsidence caused by groundwater extraction, and the loss of coastal mangroves to make way for agriculture and aquaculture production sites [16].

Apart from sedimentation fluxes, human activities have emerged as significant drivers of riverbank and coastal erosion in the VMD [17]. Notably, dams built for hydropower and sand mining along the main channels have decreased sediment availability in the Mekong River basin [18]. The presence of 56 hydroelectric plants on the Mekong River system, primarily located in Chinese territory, further exacerbates reduced sedimentation [19]. The **aowan and Nuozhadu reservoirs, completed in 2012 and 2016, respectively, have significantly impacted sediment supply to the lower stretches of the Mekong River [20].

Along the coast, there is a total of almost 50,000 ha of mangroves. Presently, the mangrove belt between the sea and the land is diminishing from both sides due to erosion, with an annual loss estimated at around 430 hectares [21]. The role of mangroves in supporting a healthy coastal environment is significant, as they serve as a natural habitat for diverse species and contribute to stable shorelines and forestry production, yet their thin belt is becoming increasingly vulnerable due to a lack of optimal protection measures and clearing for human purposes [22,23]. Many areas in the delta are experiencing coastal erosion. The coastal area of the VMD is also susceptible to the impacts of global climate change, including sea level rise and the anticipated increase in the intensity and frequency of storms and floods. Erosion along the VMD coast, spanning 245 km, has been officially documented, prompting the placement of structural measures to mitigate the widening gaps in the mangrove belt and protect the coastline [24]. Breakwaters, including semi-circular, Busadco, and pile–rock designs, have been constructed in recent years [25,26]. Furthermore, physical models of small-scale breakwaters have been studied, along with experiment models. These studies assume that the wave energy arriving at the breakwater model will be equal to the transmitted wave energy plus the energy absorbed by the breakwater and the reflected wave energy [24,27,28].

Numerous studies have been conducted to assess shoreline changes and the impacts of coastal structures on coastal evolution worldwide. In their study, Molina et al. (2019) [29] evaluated the effects of coastal structures on coastal changes in Andalusia, Spain. Their findings indicate a negative net balance of 29,738.4 m²/year, resulting in the loss of 1784.30 km² of beach surface from 1956 to 2016.

The analysis showed that rigid structures had an impact, with accretion observed primarily updrift of ports and groynes, as well as near protection structures like breakwaters.

Laksono et al. (2022) [30] examined medium-term shoreline changes in Augusta Bay, eastern Sicily, Italy, from 1972 to 2021, and evaluated the effectiveness of coastal armoring in local coastal modifications. The findings revealed a notable erosion of the shoreline in this coastal region, especially near river deltas, as a result of human and natural pressures. A study conducted by Alberico et al. (2022) [31] focused on coastal changes in the Calabria Region. They utilized satellite, drone, and field survey techniques to analyze the formation and dismantling of a fan delta at the mouth of the Sfalassà Stream.

The Prime Minister’s Decision No. 379/QD-TTg on approving the National Strategy for Natural Disaster Prevention and Control emphasizes the importance of addressing riverbank and coastal erosion, especially in the VMD region [32]. Nevertheless, the coastal provinces of the VMD have limited understanding of the root causes of current-day shoreline loss and the effectiveness of structural mitigation measures. Without such understanding, the government will continue to waste considerable public funds on ineffective solutions. This failure stems from a lack of integrated analysis of the multiple and interlinked factors causing the coastline changes. These factors include long-term sediment depletion, sea level rise, and agricultural land use practices. Quantifying the scale and speed of coastal erosion is necessary to assess land loss vulnerability in the area, along with determining the mechanical relationship between erosion mechanisms, natural factors, and human activities. In the VMD, a few studies have applied remote sensing and GIS to detect changes in the shoreline. Even less in situ research has been conducted to evaluate the effectiveness of breakwaters in reducing wave energy and coastal erosion. To develop a response or adaptation plan for development, it is necessary to identify the exact location-specific causes and mechanisms involved in coastal erosion. Monitoring shoreline changes can determine changes in shoreline location, extent, and quality, and therefore, plays a crucial role in shoreline protection. This paper aims to evaluates the current status of shoreline erosion and protection and the efficiency of various wave-reducing breakwaters in the VMD. Landsat satellite images were used to establish a coastal development map for the period 2000 to 2022. Wave data from in front and behind the breakwaters were measured and analyzed to evaluate the wave reduction efficiency of different breakwater structures.

The VMD underwent rapid progradation in a sheltered bay in the East Sea, transitioning from an estuary to a delta due to intense fluvial sedimentation [13,33,34]. Due to the increasing exposure to ocean waves during delta progradation, numerous beach ridges have formed in the eastern part, which is divided into 6–7 distributary mouths [35,36]. In the eastern area, distributary mouths are located and long-term delta evolution dominated by sandy beach ridge deposition has occurred. This significantly contrasts to the rest of the delta, where muddy progradation dominates. From a general long-term sediment-budget point of view, fluvial mud was transported southwestward by waves and currents [37,38]. Additionally, the elevation of the VMD is typically below +5 m (MSL), and it is bordered by Pleistocene uplands, swamps, and the Saigon River system to the north and northeast. On the west and southwest, it is bounded by the Ca Mau Peninsula and the Gulf of Thailand [13,39]. When assessing the sedimentation and morphology of the subaqueous layers in the VMD, according to [40], the subaqueous VMD area can be divided into five sub-areas based on clinoform morphology, influenced primarily by waves, tides, and river impact. These factors also form the basis of delta classification schemes.

2. Materials and Methods

Evaluating the effectiveness of existing protection strategies in integrated coastal management and planning involves three research components: document review, field observations, and data from annual measurement projects (Figure 2).

2.1. Data Collection

2.1.1. Landsat Images Collection and Processing

Different kinds of satellite data have been utilized for remote sensing purposes. Freely accessible Landsat satellite data, with a 30-m resolution, have been found to be particularly suitable for the extensive VMD. Landsat 5, 8, and 9 satellite imagery, spanning the period 2000 to 2022, were used to monitor the shoreline changes of the VMD. A comprehensive overview of the image particulars is encapsulated in

Table 1. Overview of utilized Landsat satellite images.

Date	Path/Row	Satellite	Sensors	Resolution (m × m/Pixel)	Ratio of Clouds (%)	Coordinates
03/03/2000	125/053	Landsat 5	TM	30	30.0	UTM
04/04/2000	125/054	Landsat 5	TM	30	30.0	UTM
10/03/2000	126/053	Landsat 5	TM	30	14.0	UTM
26/03/2000	126/054	Landsat 5	TM	30	52.0	UTM
04/05/2005	125/053	Landsat 5	TM	30	20.0	UTM
21/06/2005	125/054	Landsat 5	TM	30	18.0	UTM
19/01/2005	126/053	Landsat 5	TM	30	1.0	UTM
15/08/2005	126/054	Landsat 5	TM	30	6.0	UTM
18/05/2010	125/053	Landsat 5	TM	30	11.0	UTM
27/02/2010	125/054	Landsat 5	TM	30	29.0	UTM
09/05/2010	126/053	Landsat 5	TM	30	22.0	UTM
09/05/2010	126/054	Landsat 5	TM	30	23.0	UTM
16/05/2015	125/053	Landsat 8	OLI/TIRS	30	33.1	UTM
16/05/2015	125/054	Landsat 8	OLI/TIRS	30	13.9	UTM
21/04/2015	126/053	Landsat 8	OLI/TIRS	30	0.2	UTM
20/03/2015	126/054	Landsat 8	OLI/TIRS	30	2.2	UTM
23/02/2020	125/053	Landsat 8	OLI/TIRS	30	0.3	UTM
23/02/2020	125/054	Landsat 8	OLI/TIRS	30	1.5	UTM
13/01/2020	126/053	Landsat 8	OLI/TIRS	30	12.9	UTM
17/03/2020	126/054	Landsat 8	OLI/TIRS	30	1.4	UTM
04/02/2022	125/053	Landsat 9	OLI/TIRS	30	12.9	UTM
11/05/2022	125/054	Landsat 9	OLI/TIRS	30	52.2	UTM
18/01/2022	126/053	Landsat 8	OLI/TIRS	30	2.8	UTM
18/01/2022	126/054	Landsat 8	OLI/TIRS	30	0.5	UTM

. Following acquisition, the original Landsat images underwent geometric correction conforming to the global geographic coordinate system UTM/WGS-84, specifically within the 48N zone. This meticulous preprocessing aimed to address discrepancies in location and variations in topography.

Within the context of this study, the analytical focus predominantly revolved around shoreline interpretation, given the precedent geometric corrections. In order to facilitate a nuanced assessment, the analysis mandates the execution of image channeling procedures corresponding to distinctive Red–Green–Blue (RGB) color combinations for each variant of Landsat imagery, as delineated in

Table 2. Several RGB color combinations are used in classification.

Application Classification	Spectrum	Combination of Channels
		Landsat 5	Landsat 8, 9
Natural colors	RED, GREEN, BLUE	3 2 1	4 3 2
Vegetation (Infrared colors)	NIR, RED, GREEN	4 3 2	5 4 3
Agricultural land	SWIR-1, NIR, BLUE	5 4 1	6 5 2
Land/water	NIR, SWIR-1, RED	4 5 3	5 6 4

. This methodological approach underscored the precision and accuracy in delineating coastal dynamics within the specified geographical regions.

Classification in the realm of remote sensing pertains to the assignment of pixels into distinct classes, wherein pixels within the same class exhibit analogous properties [41,42], maximizing the probability of accurate categorization [43]. Various methodologies exist for classifying remote sensing data, with a prevalent dichotomy between supervised and unsupervised classification methods. In this study, a hybrid approach was adopted, integrating both supervised and unsupervised classification techniques. This synergistic combination was employed to optimize object extraction, yielding results of the utmost satisfaction. Specifically, the normalized difference water index (NDWI) served as a pivotal tool for the coastline interpretation of the Landsat imagery, boasting an accuracy rate of up to 90.48% [44]. The utilization of NDWI is guided by a prescribed formula [44]:

$$NDWI=\frac{GREEN-NIR}{GREEN+NIR}$$

(1)

where GREEN is the green light channel, which has a wavelength from 0.52 to 0.60 µm, and NIR is the near-infrared channel, which has a wavelength from 0.76 to 0.90 µm. For Landsat 5 satellite images, GREEN is channel 2 and NIR is channel 4 [45]:

$$NDWI=\frac{Chanel2-Chanel4}{Chanel2+Chanel4}$$

(2)

In the context of Landsat 8 and 9 satellite imagery, it is imperative to note that the GREEN component corresponds to channel 3, while the NIR component aligns with channel 5 [46]. Therefore, the formulation for the NDWI specific to Landsat 8 and 9 images is articulated as follows:

$$NDWI=\frac{Chanel3-Chanel5}{Chanel3+Chanel5}$$

(3)

This distinction is crucial for the accurate computation of NDWI, as the channels associated with GREEN and NIR differ between Landsat satellite versions. Adhering to these channel specifications ensures precision in the interpretation of coastal features and reinforces the reliability of the classification process.

Ultimately, the determined shoreline positions serve as foundational data for calculating erosion and deposition rates. The positional accuracy of the shorelines extracted from Landsat images was thoroughly investigated and discussed in [47].

2.1.2. Calculation of Shoreline Change Rate

The assessment of shoreline change rates stands as a prevalent indicator in coastal studies, denoting alterations in the shoreline position perpendicular to a designated cross-section over time [48,49]. This transformation can manifest as either long term, spanning decades or centuries, or short term, reflecting seasonal or even more rapid changes in shoreline position. Linear regression (LRR) represents a widely adopted method for quantifying shoreline change rates [50].

LRR distinguishes itself by leveraging the statistical principle of utilizing the entirety of available data points within the shoreline position data series. The methodology entails determining the slope of the regression line, serving as the numerical representation of shoreline change velocity. The formula underpinning the regression line is expressed as follows [50]:

$$y=a.t+b$$

(4)

where y represents the shoreline position (m), a is the slope of the regression line and is the rate of change of shoreline position (m/year), t is the time (year), and b is the intersection between the regression line and the vertical axis.

In this study, shoreline sections were segmented at a distance of 50 m from Tien Giang to Kien Giang to measure the shoreline position and calculate shoreline change rate, as outlined. The shoreline position in 2000 was chosen as the baseline when calculating the shoreline change rate. The value of the LRR can be negative or positive, where a positive value represents movement away from shore (deposition) and a negative value represents movement towards land (erosion). Based on the geomorphological characteristics and the rate of change of the shoreline, statistics in the study area combined with the results of the research groups [51], the average displacement rate calculated based on the LRR method was divided into 9 ranges of values based on 3 main characteristics: erosion, stability, and deposition. Each range corresponds to a hierarchy of the erosion status of the area. In particular, the erosion phenomenon was divided into 4 more detailed value ranges, including extremely high, high, medium, and low erosion. At the same time, the deposition phenomenon was also divided into 4 ranges including low, medium, high, and extremely high deposition zone.

2.2. Field Measurement

2.2.1. Wave Measurement in Front and Behind Breakwaters

Levelogger 5 Junior sensors manufactured by Solinst (Made in Canada) and INFINITY-WH AWH-USB sensors from JFE Advantech (Made in Japan) were utilized for continuous monitoring of water level, enabling the calculation of wave height over time. These sensors were securely affixed to melaleuca trees and positioned at two distinct locations: (1) 20 m seaward from the breakwaters and (2) 10 m landward from the breakwaters, as depicted in Figure 3. The sensors were strategically positioned at an elevation of approximately 1.0 m above the seabed. Wave measurements at three designated locations where breakwaters had been previously installed were undertaken (Figure 3):

• Location A: Busadco, centrifugal pile–rock and semi-circular breakwaters at Hon Da Bac, Ca Mau;
• Location B: Coastal area with and without mangrove belt, centrifugal pile–rock breakwaters at Vinh Chau, Soc Trang;
• Location C: Busaco and centrifugal pile–rock breakwaters at Ngoc Hien, Ca Mau.

Water level data were recorded at an interval of 0.5 s and converted to water column height, from which the wave height was calculated [52]. The author referred to the data processing process that has been programmed into functions in MATLAB and R programming languages [53].

2.2.2. Evaluating the Effectiveness of Wave Reduction

The wave reduction efficiency was evaluated according to two criteria: (1) reducing the height and (2) reducing the wave energy transmitted through the breakwaters. The height of waves in the area behind the breakwaters, symbol H_t, was determined using the following formula:

$$H_t=K_{tr}\times H_{sp}$$

(5)

where H_sp is the height of waves in front of the structure (m), and K_tr is the wave transmission coefficient. K_tr depends on the distance from the top of the structure to the design to sea level (h_c) and the wave height in front of the structure (H_sp). The wave reduction efficiency was calculated using the following formula:

$$\epsilon =\left(1-K_{tr}\right)\times 100\%$$

(6)

Considering cases for calculating wave reduction based on actual measurement results for three cases: (a) average 1/10 of the maximum wave height (1/10 H_max); (b) average 1/3 of the maximum wave height (1/3 H_max); and (c) average wave height (H_max).

2.2.3. Deposition/Erosion Measurement

The breakwaters were placed at localities subjected to extreme erosion (Figure 3). To determine the effect of creating alluvial ground of the breakwater at Location B with pile–rock breakwaters at Vinh Chau, Soc Trang, the author measured the elevation of 4 cross-sections perpendicular to the breakwater, as shown in Figure 4.

Sediment traps, which were rectangular with an open top, were used to measure the erosion and deposition. The size of the traps was 20 cm high with a bottom of 30 cm in length and width, and they were made of stainless steel. In order to limit the settlement of these traps over time, which would affect the measurement results, 2 bamboo trees about 2.5 m long were used to fix the 2 sides of the trap with the sand layer below. The measurement was carried out at four sections along the coast with the interval of 300 m. Each section had 3 measuring points: X and Y are far from the breakwater, 100 and 50 m to the shore, respectively and Z is 30 m far from the breakwater toward the sea. The measurement process was carried out periodically once a month using a hydrometer to determine the change in the height of the cross-sections over time. The elevation landmark was taken as the level of sluice gate No. 2 (+4.130 m).

2.2.4. Monitoring Settlement of Breakwaters

Settlement observations were carried out on two components of the pile–rock breakwater at Location B: (1) the concrete part on the top of the breakwater, and (2) the rock layer located between the concrete and the pile. The monitoring points were marked on the breakwater at the interval of 300 m along the breakwater route; they are shown on the ground in Figure 5. Because the surface of the rock was not flat, we undertook a fixed measurement at a distance of about 50 cm from the observation point of the concrete settlement. The observation period started on 14 October 2021 and ended on 17 July 2022.

The settlement of the rock part consisted of two factors: (1) the consolidation of the ground below the rock layer, and (2) the expected rearrangement and settlement after a period of operation under the influence of sea waves. The calculated settlement of the ground under the rock was calculated using the layer-by-layer settlement method according to Vietnamese construction standards (TCVN 4253:2012).

2.3. Breakwater Structure Stability Analysis

The breakwater structures at Location A were utilized for stability assessments at two distinct phases: (1) following their construction in 2018 and (2) after five years of operational use in 2023. Operational observations highlighted the efficacy of sedimentation creation behind the breakwaters. However, during the operational period, vortexes developed at the base of the breakwaters, resulting in reduced ground elevation at the foundations of the breakwaters. We collected data on deposition and erosion for various breakwater types (

Table 3. Elevation measurement data of the breakwaters area.

Elevation Measurement Periods	Distance from the Breakwater to the Shoreline
	Busadco Breakwater					Pile–Rock Breakwater					Semi-Circular Breakwater
	1 m	5 m	10 m	15 m	20 m	1 m	5 m	10 m	15 m	20 m	1 m	5 m	10 m	15 m	20 m
Construction Completed (2018)	−0.95	−0.9	−0.85	−0.9	−0.95	−0.92	−1.05	−0.95	−1.05	−1.02	−0.9	−0.85	−0.9	−1	−1.05
12 May 2023	−1.2	−0.27	−0.24	−0.25	−0.26	−1.19	−0.17	−0.21	−0.22	−0.2	−1.2	−0.24	−0.25	−0.23	−0.27
Change of elevation	−0.25	0.63	0.61	0.65	0.69	−0.27	0.88	0.74	0.83	0.82	−0.3	0.61	0.65	0.77	0.78

), employing these data as input for stability calculations using the finite element method (FEM). The computational scenarios encompassed the lowest and highest sea levels corresponding to the aforementioned stages.

The stability of these structures has not been investigated in the literature [11,24,25,26]. After declaring the geometric, geological, water level, and wave pressure characteristics of the structure, we calculated the Factor of Stability (FS) using the method that involves reducing shear resistance on the FEM. This method is widely utilized in studies, and it is formulated as follows [54,55,56]:

$$FS={\displaystyle \sum M_{sf}}=\frac{\tan \phi }{\tan {\phi }_r}=\frac{c}{c_r}$$

(7)

where c is the cohesive strength in the normal state (kN/m²), φ is the internal friction angle in the normal state (°), c_r is the reduced cohesive strength sufficient to maintain equilibrium (kN/m²), and φ_r is the reduced internal friction angle sufficient to maintain equilibrium (°).

The reduction in parameters c and φ is governed by the variable ΣM_sf, which varies at each calculation step until instability occurs. At this point, the calculation stops. The final value of ΣM_sf that is obtained just before instability is the FS. In the method of reducing shear resistance:

• FS > 1 indicates that the structure is in a stable state;
• FS = 1 implies the structure is at the limit of instability;
• FS < 1 indicates that the structure is in an unstable state, posing a high risk of sliding.

In this study, PLAXIS 2D 2021 FEM-based software was used to assess the stability of the breakwaters. This FEM software has been significantly used for finite element analysis since 2010 [57]. Additionally, the field-recorded data, used as the input for PLAXIS 2D 2021 FEM, are presented in

Table 4. Data on soil mechanical characteristics.

No.	Soil Mechanical Characteristics	Symbols	Unit	Layer 1: Clay Silt	Layer 2: Mixed Clay
1	Natural Moisture Content	w	%	75.86	29.96
2	Unsaturation Unit Weight	γunsat	kN/m3	15.72	19.71
3	Saturation Unit Weight	γsat	kN/m3	15.96	19.97
4	Specific Gravity	Gs	-	2.654	2.707
5	Internal Friction Angle	φ	°	3.51	14.19
6	Cohesion Strength	C	kN/m2	6.67	26.18
7	Stiffness Module	E	kN/m2	598	5815
8	Hydraulic Conductivity	K	m/day	5.22 × 10−3	4.27 × 10−3

.

During the computational process, the author considered the impact of the breakwater structures on deposition and erosion, especially focusing on the observed phenomena of whirlpools affecting the ground elevation at the seawall base. The FEM-based stability analysis aimed to provide insights into the performance and potential challenges associated with the breakwater structures, considering both the design and operational phases. Data on soil mechanical characteristics are presented in Table 4.

3. Results and Discussion

3.1. Shoreline Change

Figure 6 depicts the shoreline positions from 2000 to 2022, revealing a dynamic pattern of deposition and erosion along the coast of the VMD. Notably, the northern coastal provinces, such as Tien Giang, Ben Tre, Tra Vinh, and Kien Giang, exhibit a juxtaposition of deposition and erosion events. In contrast, the southern coastal provinces, such as Soc Trang, Bac Lieu, and Ca Mau, predominantly experienced erosion along many of their coastlines during this period.

Table 5. Length and rate of erosion/deposition in coastal provinces of VMD.

Province	Coastal Length (km)	Erosion			Deposition
		Erosion Length (km)	Percentage	Erosion Rate (m/Year)	Deposition Length (km)	Percentage	Deposition Rate (m/Year)
Tien Giang	33.35	21.30	63.9%	3.03–18.91	3.95	11.8%	3.02–8.42
Ben Tre	88.10	25.85	29.3%	3.01–36.75	29.15	33.1%	3.00–71.62
Tra Vinh	73.25	16.35	22.3%	3.02–20.92	34.70	47.4%	3.03–38.51
Soc Trang	85.60	14.75	17.2%	3.05–15.41	54.70	63.9%	3.08–57.61
Bac Lieu	54.75	28.70	52.4%	3.03–15.48	12.00	21.9%	3.02–28.20
Ca Mau	275.30	165.25	60.0%	3.01–67.35	61.55	22.4%	3.03–94.18
Kien Giang	214.95	23.15	10.8%	3.02–18.95	69.30	32.2%	3.00–73.87
Total	744.0	295.35	-	-	265.35	-	-

presents an overview of the total length of erosion and deposition sections from 2000 to 2022, accompanied by detailed rates of accretion and erosion for various provinces. Ca Mau, with the longest coastline, exhibits the highest erosion rate among the coastal provinces, reaching 60.0%. The pace of coastal erosion in Ca Mau is notably swift, with a section eroding at a rate of 67.35 m per year. Specifically, two locations in Ca Mau, labeled A and C (refer to Figure 7a,c), demonstrate exceptionally high erosion rates, with Location A experiencing up to 27.7 m/year and Location C reaching 35.9 m/year. These elevated erosion rates can be attributed to the comparatively recent formation history of the Ca Mau peninsula in contrast to other provinces of the VMD.

The shoreline to the south of this area exhibited a higher rate of erosion (up to 15.5 m/year) than the north, where there are alternating sections of erosion and deposition (Figure 7b). Locations B and C were influenced by the semi-diurnal tide regime of the East Sea when the high and low tide repeats two cycles in a day combined with strong northeast monsoons in the months from October to March [24], which led to extremely forceful coastal erosion.

3.2. Coastal Protection Solutions

Mangrove belts have protected the coastal area of the VMD since its formation, with their roots also contributing to strengthening the ground and reducing the impact of waves on the land behind. However, during the last two decades, the area of coastal mangroves has seriously decreased in provinces, such as Soc Trang, Bac Lieu, and Ca Mau. Protecting these forests requires new coastal protection solutions that will not rely solely on nature. In the coastal provinces of the VMD, many structures to protect the shore have been built, such as sea dikes, revetments, centrifugal pile–rock breakwaters, Busadco breakwaters, semi-circular breakwaters, and bamboo T-fences. These are outlined in Table 5 below. The above structures are preliminary effective at lowering the rate of erosion and storing the amount of sediment required to create alluvial ground. The research team carried out field surveys at construction sites in locations A, B, and C during this study (Figure 6 and Figure 8). According to [24], up to 199.4 km of shoreline has revetment or breakwaters. This percentage is approximately 68%, showing that the coastal province governments are aware of the danger of the current erosion situation and are responding.

3.3. Wave Height and Wave Reduction Analysis

3.3.1. Wave Characteristics

Location A, on the west coast of the VMD, experiences the full impact of the southwest monsoon, prevailing during the rainy season from May to November each year. To assess the wave dynamics in this region, the research team conducted wave height measurements both in front of and behind the Busadco breakwater on 29 June 2019. Subsequently, the team extended these measurements to encompass three different types of breakwaters in this location (pile–rock, Busadco, and semi-circular breakwaters) on 15 October 2020. The outcomes of the wave composition before and after the implementation of these three types of breakwaters are illustrated in Figure 9 and Figure 10.

The wave characteristics at the two points (in front and behind the pile–rock breakwaters) are depicted in Figure 11 and Figure 12. Figure 11 depicts the wave measurement results in the dry season (January), while Figure 12 shows the wave measurement results in the rainy season (July). It can be seen that the wave height in the July measurement is higher than that of January.

At location C, the research team conducted wave height measurements before and after the implementation of the Busadco breakwater on 7 April 2019. The findings, including water level heights and wave composition both before and after the installation of the breakwater, are presented in Figure 13.

The research team conducted wave height measurements both before and after the installation of the pile–rock breakwater on 23 July 2022 and on 18 March 2023. The outcomes, depicting water level heights and wave composition before and after the implementation of this breakwater, are visualized in Figure 14 and Figure 15.

3.3.2. Wave Reduction Effect

Based on the outcomes of the wave height measurements over time and the analysis of wave height components, the research team quantified the impact of wave height reduction according to the following criteria: maximum wave height (1/10 H_max), significant wave height (1/3 H_max), and average wave height (H_max). The calculated results are depicted in Figure 16, providing insights into the effectiveness of wave height reduction achieved by the implemented coastal protection measures.

Based on the analysis of wave height data over time and the calculation of wave reduction efficiency, a correlation was observed between the efficiency of different breakwater types and the recorded maximum wave height. The Busadco breakwater, characterized by a hollow triangular structure, situated at Location A, exhibited the highest efficiency, achieving approximately a 90% reduction in both significant wave height and average wave height. Conversely, the pile–rock breakwater at Location C displayed the lowest wave reduction efficiency, primarily due to the absence of recorded highest wave height data in this area. Furthermore, being the inaugural pilot breakwater in the VMD, the design of the pile–rock breakwater may not be optimized for wave reduction efficiency.

3.4. Depostion Analysis

The research team systematically collected sediment samples at two designated areas for deposition/erosion measurement monitoring (both in front of and behind the centrifugal pile–rock breakwater) at Location B. Sampling intervals were established approximately once every 2–4 months on the following specific dates: 13 October 2021, 21 January 2021, 15 May 2022, and 17 July 2022. Grain distribution curves representing the sediment at the two positions, i.e., in front of and behind the breakwater, are depicted in Figure 17. Utilizing the distribution curves, the average particle diameter (d50) was calculated and is presented in

Table 6. The average particle diameter (d50).

Collection Date	d50 in Front Breakwater (mm)	d50 behind Breakwater (mm)
13 October 2021	0.016	0.001
15 January 2022	0.019	0.002
15 May 2022	0.023	0.004
17 July 2022	0.034	0.006

.

The observed difference in average particle size between the front and rear of the breakwater signifies a reduction in particle size towards the rear, indicating the movement of fine particles from the front to the rear. This trend suggests effective sediment deposition by the breakwater, capturing and retaining fine particles inside.

The deposition/erosion rates in the vicinity of the pile–rock breakwater are illustrated in Figure 18, with measurements initiated on 15 April 2020, shortly after the completion and commencement of the breakwater’s operation, until 17 July 2022. Notably, positions from the breakwater back to the mainland exhibit a substantial increase in elevation. At cross-sections 1, 2, and 3, marked elevation changes occurred at points X and Y (100 m and 50 m from the breakwater to the mainland, respectively), ranging from 0.4 to 0.9 m. Conversely, at position 4Z (30 m away from the breakwater towards the sea) outside the breakwater’s protective area, no significant deposition was observed, with slight erosion occurring.

In conjunction with the particle composition test results, the research team deduced that fine particles from outside the breakwater (at point Z) migrated towards points X and Y, initiating a process of deposition within the breakwater and erosion outside its protective barrier.

The average deposition rates within the breakwater, observed over a measurement period of 27 months (from 15 April 2020 to 17 July 2022), reveal varying rates at positions X and Y of cross-sections 1, 2, and 3. Specifically, the deposition rates were measured at 3.1, 2.9, and 2.0 cm/months, respectively. Notably, at cross-section 4, the average deposition rate is −0.8 cm/month, indicating slight erosion processes. These findings are illustrated in Figure 19. The deposition predominantly occurred within the breakwater at cross-sections 1, 2, and 3, while outside the breakwater area at cross-Section 4, the deposition process was minimal, accompanied by slight erosion.

It is important to note that the coastlines of the VMD are influenced by seasonal wave climates driven by the northeast and southwest monsoons. During the northeast monsoon season, which lasts from November to March, the primary wave direction along the east coast of the Mekong Delta is from the northeast, bringing the highest waves of the year during winter. In contrast, during the southwest monsoon season, from May to September, waves predominantly approach both the east and west coasts of the Mekong Delta from the southwest [58].

Albers et al. [59] conducted extensive field measurements along the coastlines of Soc Trang and Bac Lieu provinces through three separate campaigns. These campaigns recorded data on currents, waves, sediment concentrations, and bathymetry. Their findings indicated a clear dependency of wave measurements on the monsoon seasons. The recorded currents exhibited a strong longshore component, intensified by the northeast monsoon due to the tidal wave along the South Vietnamese coast. During the peak period of the northeast monsoon, higher waves were observed in the focus area, approaching the coasts of Soc Trang and Bac Lieu with a pronounced longshore component. In winter, although the sediment plume of the Mekong is less pronounced and less material is available, the northeast monsoon winds result in increased coastal longshore drift and erosion [60].

Figure 18 shows a section of the breakwater placed perpendicular to the shoreline (near sluice gate number 2). This part of the breakwater acts as a groyne, blocking longshore sediment transport in the northeast direction and causing sediment accumulation on the updrift side of the groyne. This phenomenon is similar to the findings of [61,62]. Therefore, the highest amount of sediment deposition was observed at location 1 (Figure 19), closest to the perpendicular part of the breakwater (Figure 18). The amount of sediment deposition decreased from location 1 to location 3, with even erosion observed at location 4 (Figure 19). This indicates that longshore sediment transport is dominant in the northeast direction, which is consistent with the studies of [60].

3.5. Settlement of Pile–Rock Breakwater Components

During the first two months of monitoring the breakwater settlement at Location B (from 14 October 2021 to 23 December 2022), the settlement of the concrete components of the breakwater rose by an average of 7.0 to 8.0 mm/month. In the remaining 7 months (from 23 December 2021 to 17 July 2022), the consolidation process was close to zero. In general, after more than nine months of monitoring the maximum settlement of the breakwater is at point 2 with −21 mm, this value is very small compared to the limited settlement of hydraulic structures. The above results show that the stability of the centrifugal pile–rock breakwater is very high (Figure 20a).

The results of monitoring the settlement of the rock in the middle of the concrete components of the breakwaters are shown in Figure 20b, showing that the largest settlement was −290 mm at point 3 after more than nine months of monitoring. Compared to the concrete components, the settlement of the rock pit was relatively large. The settlement of the pit rock is the sum of (1) the settlement of the soil layer below and (2) the settlement related to the rearrangement of the rocks. The results of calculating the settlement of the soil according to the Vietnamese foundation design standard (TCVN 10304:2012) show that the settlement of the soil was up to −416 mm. Combined with the decrease in the settlement rate of the rock elements, it can be seen that the consolidation settlement is nearly complete.

3.6. Stability Analysis

The method of diminishing shear resistance implemented in PLAXIS 2021 FEM software was employed to compute the comprehensive stability factor of breakwaters at geological borehole positions along the breakwater alignment. The outcomes encompass two scenarios corresponding to the highest and lowest sea levels, as depicted in Figure 21. Aligned with the moderately weak geological structure of the VMD extending to a depth beyond −15.0 m, geological parameters such as cohesive strength (c = 6.67 kN/m²) and internal friction angle (φ = 3.51°) govern the structure’s stability.

The stability calculation results shown in Figure 21a, considering the scenario with the highest sea level, demonstrate increased stability coefficients for all types of embankments after 5 years of operation. Conversely, in the case of the lowest sea level (Figure 21b), it is noted that the stability coefficients for the Busadco embankment and the eccentric embankment decline over time. Particularly, the Busadco embankment experienced a reduction from 2.14 to 1.68. Meanwhile, the stability coefficients for the hollow cylindrical embankment were augmented, akin to Figure 21a. This can be explained by the geometric structure of the crescent embankment, which redirects external forces toward the embankment’s center. In contrast, the triangular hollow structure of the Busadco embankment, which is subject to pressure from one side along with vortex holes at the embankment’s base, exhibits a heightened susceptibility to instability.

These findings substantiate the compliance of the breakwaters’ stability factor with Vietnam construction standard 04-05:2012/MARD [63], pertaining to the design stability factor required for hydraulic structures.

4. Discussion

Breakwaters, mangrove restoration, and the utilization of remote sensing applications in shoreline monitoring can be widely applied to other deltas and coastal areas around the world that confront similar long-term sustainability challenges driven by compounding anthropogenic and environmental issues, such as sediment reduction, land subsidence, sea level rise, and heightened storm intensity. Studies focusing on the Nile Delta, the Ganges–Brahmaputra Delta, the Indus River Delta, and the Mississippi Delta have all reported similar challenges [64,65,66,67,68,69].

Our findings are based on a limited dataset, which included wave heights predominantly below 1 m. In future studies, we hope to incorporate additional water data from a more extended time period and from a more comprehensive dataset. Nevertheless, our findings offer insights into the effectiveness of different coastal protection mitigation measures and, as such, provide critical guidance for coastal planners in the VMD and for regions implementing similar interventions.

Furthermore, our results emphasize the necessity of develo** comprehensive assessment frameworks that include socio-economic and environmental impacts to guide integrated coastal management practices. Policymakers increasingly need to take note of the importance of sustainable coastal management, combining hard, hybrid, and soft engineering solutions, the ecological and social considerations, policy adjustments, and community involvement. Although the fundamental principles and methods of our study can be applied elsewhere, the specific solutions need to be adjusted to fit the local environmental, social, and economic conditions. The unique characteristics of each region, such as wave energy, sediment type, and ecological settings, must be considered. Overall, this study provides a robust and adaptable framework for monitoring coastal erosion in diverse global contexts, contributing to broader coastal resilience and sustainability. Future work should involve the collection and analysis of more comprehensive wave dataset to build upon our findings.

5. Conclusions

Coastal environments at low elevations are highly susceptible to both natural and human-induced disturbances. In the VMD, coastal erosion has become a critical issue, affecting 295.4 km of the 744.0 km coastline, with a significant portion experiencing severe erosion. The erosion rate in the southern provinces of the VMD has significantly increased over the past two decades, reaching an alarming rate of nearly 95 m per year. In response, most coastal provinces have implemented protection measures, now covering 68% of the eroded coastline. Breakwaters have proven highly effective, reducing wave height by over 62% and significantly mitigating the impact of waves on the shoreline.

Our observations have shown that the creation of offsets from the breakwater to the mainland, with compensation rates increasing at a rapid pace of 3.1 cm/month. Despite their high stability, pile–rock structures still require periodic rock replenishment to counter settlement in the rock sections. The importance of ongoing maintenance in coastal environments is highlighted by the need for rock additions to maintain the effectiveness of protective structures.