Quantifying Spatiotemporal Evolution of Sandy Shorelines in Northern China Using DSAS: A Case Study from Dalian World Peace Park

Abstract

This study analyzed shoreline evolution (2000–2024) at Dalian World Peace Park’s sandy tourist beach using GEE, CoastSat, and DSAS. At the same time, combined with the grain size analysis of beach sediments before and after typhoons, the impact of extreme events on the shoreline line changes was explored. The DSAS shows a spatial differentiation pattern of the southern shoreline retreat trend zone, the central shoreline dynamic balance trend zone and the northern shoreline advance trend zone. The 2008 reclamation project altered hydrodynamics, creating an artificial headland effect that triggered significant northern shoreline advancement (max 74.16 m) and southern retreat (27.14 m), demonstrating unforeseen long-term trade-offs of large-scale interventions. Subsequent cobble structures, acting as a nature-based solution, enhanced sediment retention and wave energy refraction, promoting dynamic equilibrium and shoreline resilience. However, the 2017 double typhoon caused instantaneous retreat with finer, poorly sorted sediment, highlighting persistent vulnerability to extreme events. This study underscores the critical need for adaptive management within a sustainable shoreline development framework.

1. Introduction

The coastal zone, where the sea, land, and air interact, possesses significant ecological, economic, and social value. Shorelines are classified into bedrock, sandy, muddy, biological, and estuarine types according to substrate properties and spatial forms [1]. As the predominant coastal type, sandy shorelines are crucial for maintaining coastal ecosystem stability, mitigating marine disasters, and promoting tourism development, owing to their unique high permeability and strong plasticity [2]. Sea-level rise has been accelerated by global climate change, and human activities have intensified. These processes have led to a sediment supply–demand imbalance, loss of natural shoreline restoration capacity, and continuous shoreline retreat. The combined impacts of climatic and anthropogenic pressures have aggravated systematic shoreline retreat of sandy coasts [3]. Especially in the northern regions of the world, the superposition of seasonal storms, freeze–thaw cycles, and high-intensity human activities have resulted in complex deposition-dominated change patterns in sandy coasts. Statistics show that more than 70% of the global sandy coasts experience shoreline retreat, with rates generally exceeding 1 m/y and locally reaching 1.5–3.0 m/y, thus directly threatening the integrity of coastal infrastructure and ecosystem services [1,3,4]. Complex spatiotemporal fluctuations occur in northern sandy beaches due to the combined effects of winter cold fronts, summer typhoons, monsoon waves, and human activities [5,6]. Compared with pristine beaches, developed sandy coasts face greater stability challenges from increased tourist activities, continuous facility construction, and artificial restoration measures, which accelerate processes such as shoreline retreat, shoreline advancement, shoreline retreat, and morphological modification [7,8].

Traditional studies on coastline evolution primarily rely on regular on-site GPS measurements, low-resolution remote sensing imagery analysis, and comparisons of historical topographic maps and aerial photography. Centimetre-level accuracy can be achieved by ground measurement technologies such as total stations and RTK-GPS. However, extended data collection cycles and limited geographical coverage are recognized as drawbacks [8,9,10]. Interannual shoreline monitoring was enabled by Landsat and Sentinel-2 satellites. However, event-scale coastal dynamics are constrained by intrinsic temporal thresholds. The standard 5–16 d revisit cycles preclude the detection of short-term perturbations (e.g., storm-induced sediment remobilization). Concurrently, decimeter-scale morphological changes are obscured because of the 10–30 m spatial resolution. These limitations inhibit the quantification of ephemeral sedimentary processes along sandy shorelines [11,12]. Long-term evolutionary patterns can be replicated by numerical models such as Liteline and Genesis [13]. However, ambiguity exists in the description of complex boundary conditions and multifactor coupling effects, and there is a lack of high-resolution empirical data for verification [14]. Traditional interpretation techniques, including the waterline method and Digital Shoreline Analysis System (DSAS), are significantly affected by tidal fluctuations and beach humidity, resulting in interpretation inaccuracies exceeding three pixels [15]. Contemporary research focuses mainly on interannual or seasonal changes, while cumulative impacts over decadal timescales (>10 years) have often been neglected. Due to the lack of continuous observational data on decadal scales, a comprehensive analysis of progressive coastal landform processes and their underlying mechanisms is limited. Current limitations hinder understanding. Insufficient analysis of short-term drivers and weak links between shoreline change and sediment properties pose challenges. Accurate quantification of how coastal engineering works affects long-term sediment dynamics and stability is thus prevented. Existing coastal morphology simulations exhibit critical limitations, notably overlooking wave dynamics and lacking sufficient monitoring data. These constraints impede the mechanistic understanding of sediment transport processes. For instance, Florida case studies demonstrate that tidal inlet interactions compromise beach nourishment efficacy, hindering long-term project evaluation. Collectively, such knowledge gaps obstruct evidence-based beach management by preventing the resolution of key sediment-wave feedback mechanisms [16,17,18].

This study focuses on the sandy tourist beach area of Dalian World Peace Park, utilizing remote sensing, CoastSat, and DSAS analyses to investigate shoreline changes (with specific sampling intervals detailed in Section 3.1). Additionally, by combining grain size analysis of beach sediments sampled before and after typhoons (sampling timelines provided in Section 3.1), the study explores the impact of extreme events on shoreline changes. To this end, Landsat and Sentinel-2 image data were integrated using the Google Earth Engine platform [19,20]. Subsequently, subpixel edge detection combined with tidal correction established a relatively high accuracy shoreline time series, effectively overcoming the spatiotemporal limitations of traditional techniques [14]. Meanwhile, statistical methods including the Weighted Linear Regression (WLR), End Point Rate (EPR), and Net Shoreline Movement (NSM) of the DSAS were applied to reveal the impacts of natural processes and human activities on shoreline evolution. Critically, the multi-timescale coupling analysis methodology adopted in this study successfully overcame the inherent limitations of traditional single-time series research, thus enabling a systematic analysis of beach evolution characteristics and their driving mechanisms from multiple temporal dimensions. These mechanisms encompass multiple factors, including natural processes, anthropogenic activities, and disaster events. The multi-timescale methodology provides critical insights for sustainable shoreline management by resolving drivers across event-to-decadal scales. Specifically, our analysis of engineering impacts (e.g., 2008 reclamation) and nature-based stabilization (cobble structures) offers empirical guidance for climate-resilient interventions. Ultimately, the research results can provide scientific decision support adaptive strategies for the control of coastline retreat, beach maintenance, and spatial planning of sandy coasts in the north.

2. Study Area

The southern coastal zone of the Liaodong Peninsula represents a typical area of sandy coasts in northern China, with a total coastline length of approximately 1500 km. Sandy coastal segments account for about 8% of this length. Their dynamic stability shows considerable variation due to the collective influence of monsoons, tides, and human activities. The study area is situated at the southern extremity of the Liaodong Peninsula in China (38°43′–40°58′ N, 121°05′–124°13′ E), adjacent to the Bohai Sea and the Yellow Sea. It is located in a moderate temperate monsoon climatic zone and has typical features of a bedrock-type sandy beach. The region’s terrain is primarily characterized by low mountains and hills. It has an average annual temperature of 9–10 °C, a frost-free period of over 180 days, and average annual precipitation of 550–900 mm, mostly in the summer [21,22].

Based on meteorological records from the Shuishiying Station, the local annual mean wind speed is 3.82 m/s, with the maximum wind speed reaching 32 m/s [23]. The dominant wind direction is NNE, with a frequency of 31%, followed by N at 27%, whereas the strongest winds are documented in the NNW direction. For the analysis of deep-water wave conditions, data from the Beihuangcheng Ocean Observatory, located approximately 92 km offshore, were employed [24]. The wave regimes encompassing the S-SW-W-NW directions recorded at this observatory have been validated to accurately represent the wave characteristics of the study area. Observational data from this station indicate that the dominant wave directions are NW and NNW, with the highest-energy waves emanating from the NW direction [25].

In winter, dominant winds blow from the north and northwest, and the region endures cold waves. The annual snowfall ranges from 300 to 600 mm. Tidal classifications in coastal waters demonstrate geographical variability, with the Bohai Sea coast displaying irregular semi-diurnal tides and the Yellow Sea coast displaying standard semi-diurnal tides. The distribution showed an upward tendency from south to north. The mean tidal range along the Laotie Shan-Fuzhou Bay coastline is 1–1.5 m, escalating to 2–2.5 m around the Liaohe River estuary, and attaining a maximum of 4–4.5 m at the Yalu River estuary. The mean tidal range fluctuates between 1.0 and 4.5 m. The Liaodong Peninsula features a coastline of approximately 1500 km, categorized into two primary categories based on shoreline retreat and deposition [21,22]. The bedrock shoreline accounts for 82% of the whole coastline, whilst sandy beaches represent around 8%. Due to the interplay of changes in coastal current sediment transport, tidal stresses, and monsoons, sandy beaches demonstrate inadequate dynamic stability and considerable geographical variation [26,27].

The research location is the beach at World Peace Park in Lvshunkou District, Dalian City (Figure 1), which sits on the southernmost tip of the Liaodong Peninsula, near the Bohai Strait. This semi-enclosed bay is characterized by a sandy beach landform unit. It extends approximately 1 km along a northwest–southeast axis, with a width of about 400 m and an average slope of roughly 0.032°. Gravel and medium-coarse sand constitute the predominant sediments. A key feature distinguishing this beach from others on the Liaodong Peninsula is its profound engineering impacts and anthropogenic modifications. In 2005, as part of initial tourism development, artificial sand enhancement was carried out. Land reclamation works were initiated in the northern sector of the study area in 2008. The project was completed in 2010, resulting in the formation of an artificial headland. This structure extended 1.3 km seaward. Consequently, wave energy distribution patterns were potentially modified across the study area. Subsequently, during a 2015 ecological restoration project, approximately 17,000 m³ of pebbles (median diameter: 2 cm) were nourished along an 800 m shoreline section oriented north–south. Through this intervention, a 5 m wide cobble beach berm was constructed. The landward and seaward elevations of this shoulder are 3.0 m and 2.7 m, respectively, and it extends seaward at a 1:9 slope to the original mudflat. The southern coastal section, approximately 200 m in length, was designated for revetment construction. The revetment utilized a vertical retaining wall structure, cast in concrete. The base was compacted with block stones to form the foundation, while the seaward face was armored with ballast blocks. The retained area behind the walls was backfilled with sand and gravel. The resulting structure has been found to increase wave energy reflection; meanwhile, through adjustments to sediment retention and redistribution, it has contributed to improving the stability of the beach berm. However, recent pressures, including elevated sea levels, extreme weather events, and concentrated tourism activities, have led to significant transformations in shoreline morphology and the sedimentary environment. Consequently, this site, which has undergone anthropogenic alteration and is subject to climatic stress, now serves as an ideal natural laboratory for investigating coastal dynamics.

Hydrodynamically dominated by monsoon-driven waves (mean height: 1.2–1.8 m) and irregular semidiurnal tides, the study area exhibits wave-tide interactions characteristic of semi-enclosed bays, showing similarities with other semi-enclosed bays, such as the Adriatic or Chesapeake, in terms of interactions between waves and tides. Sedimentary dynamics are governed by gravel and medium-coarse sand, though localized fine-grained layers from 2005 nourishment create textural heterogeneity akin to modified global tourist beaches. The coastal region of the research area was impacted by the dual typhoons Haitang and Noru in early August 2017, which led to heavy rainfall in Dalian City. The study area is characterized as a microcosm of this coastal unit, featuring a semi-enclosed bay approximately 1 km in length. Similar hydrodynamic settings are shared with adjacent tourism beaches, such as Fujiazhuang and Jinshitan. However, due to more intensive anthropogenic modifications, the area demonstrates heightened susceptibility to the coupling effects between human activities and natural processes. In contrast, annual shoreline change rates in adjacent unmodified sandy zones are generally less than 1 m/yr [1,2,3]. Within this study area, however, localized extreme disparities in shoreline advancement and retreat rates have been induced by engineering interventions. This highlights the magnified impact of anthropogenic alterations on small-scale coastal systems [28].

Synthesizing these characteristics, the Dalian World Peace Park beach, despite its geographical confinement, functions as a microcosm of the core processes that affect global coastal systems. Its temperate monsoon setting, with northwest-dominant waves and semi-diurnal tides, mirrors the conditions found on mid-latitude tourist coasts in the Mediterranean and the southeastern United States. Furthermore, its transition from natural gravel-sand mixtures to anthropogenically modified sediments reflects the alterations faced by over 60% of global sandy coasts due to urbanization [26,29]. Thus, the site encapsulates universal challenges, including engineering-induced morphological shifts, responses to extreme storms, and the complex balance between tourism development and ecological stability.

3. Data and Methods

3.1. Data Source

3.1.1. Remote Sensing Image Data

This work used Landsat 5/7/8/9 and Sentinel-2 series multispectral remote sensing images as the principal data sources for coastal monitoring, including spatial resolutions of 30 m and 10 m, spanning the period from 2000 to 2024. All images were taken in batches using the Google Earth Engine (GEE) platform and subjected to Top of Atmosphere (TOA) atmospheric correction [19,20]. Images with cloud below 20% that completely covered the coastal zone of the study area were selected for coastline extraction and time series analysis [29].

3.1.2. Auxiliary Data

This study utilizes a comprehensive methodology using multi-source data for the examination of shoreline dynamics. Tidal data were acquired from the MIKE Global Tide model to create instantaneous tidal heights, utilized for tidal correction in remote sensing-based coastline extraction. Topographic data were acquired using RTK-GPS to determine the average slope of the beach surface and elevation information, facilitating spatial translation calculations for shoreline positioning. Meteorological data were synthesized from the China Meteorological Observatory (https://typhoon.nmc.cn/web.html) (accessed on 19 August 2025), encompassing typhoon trajectories, wind velocity and precipitation records to carefully evaluate the processes via which extreme climatic events affect coastline change. Typhoon wind-wave data (Haitang and Noru) were provided by China’s National Marine Environmental Monitoring Center. Typhoon path imagery came from the Hong Kong Observatory, China.

3.1.3. Field Sampling Data

In 2017, four transects were set up on the beach in a north-to-south direction, with each transect spaced 250 metres apart, to systematically capture sediment changes before and after the typhoon. The first data collection was conducted at the low tide level on 28 July 2017, before the typhoon, and the second collection took place 72 h after the typhoon. During each sampling, surface sediment samples (500 g each) were collected from the high tide flat, middle tide flat, and low tide flat of each transect at a depth of 0–5 cm, resulting in a total of 12 sampling points (Figure 2). Meanwhile, a DJI Mavic 3 drone equipped with an RTK module was used for on-site recording. All field data were geotagged using a handheld GPS to ensure precise spatial alignment with remote sensing datasets, facilitating direct integration with grain size analysis results and shoreline evolution models. This RTK-enabled drone provided high-precision positioning data, facilitating accurate registration and correction in subsequent data processing, thus offering reliable measurement data for grain size analysis and shoreline evolution verification.

3.2. Data Processing and Analysis Methods

3.2.1. Shoreline Extraction Method

This study used the CoastSat toolkit (https://github.com/kvos/CoastSat, accessed on 19 August 2025) on the Google Earth Engine (GEE) platform for shoreline extraction and processing. Initially, during the picture preparation phase, a cloud mask was created with the Quality Assessment (QA) band to exclude clouds, cloud shadows, and ice-covered regions, thus ensuring the integrity of the image data. Next, the Modified Normalized Difference Water Index ($MNDWI$) was calculated to effectively distinguish between water and land areas [30]. The formula of $MNDWI$ is as follows (Equation (1)):

$$MNDWI={\displaystyle \frac{Green-SWIR1}{Green+SWIR1}}$$

(1)

$Green$ and $SWIR1$ denote the reflectance of the green light band and the short-wave infrared band. The $MNDWI$ value range is (−1, 1), with aquatic regions generally displaying elevated values and terrestrial regions showing diminished values.

To improve coastline detection accuracy, the extracted land-water boundary underwent morphological filtering for noise removal. This processed boundary was then integrated with the Otsu algorithm to precisely segment transition zones exhibiting maximum variance between sand and water [31]. The approach generates a preliminary waterline, which is subsequently smoothed and filled using morphological processing to provide an accurate continuous coastline vector. To eliminate the effects of tidal differences, tidal corrections were applied to multi-phase coastlines based on tidal data provided by the MIKE Global Tide model, thereby obtaining a standardized coastline dataset. The rate of shoreline change was ultimately determined by a superposition analysis approach (WLR, EPR, and NSM).

3.2.2. Shoreline Change Analysis Methods

The fourth-generation Digital Shoreline Analysis System (DSAS v5.0), an open-source GIS extension module developed by the US Geological Survey (USGS), was used in this study to conduct a quantitative analysis of the spatiotemporal evolution characteristics of the shoreline at Dalian World Peace Park’s sandy beach [32]. According to relevant standards, the landward migration of the coastline relative to the baseline is characterized as shoreline retreat, while the seaward migration is characterized as shoreline progradation.

Topological consistency processing was implemented on multi-source coastal time-series datasets (2000–2024) using ArcGIS 10.8’s Spatial Join tool to unify coordinate systems (WGS 84 51N) and integrate topology for multi-source coastal datasets. Based on the merged shoreline vector layer, a 100 m wide coastal zone dynamic analysis area was constructed using a buffer generation algorithm. This buffer zone effectively covered the maximum shoreline oscillation amplitude during the study period. A reference baseline, parallel to the shoreline’s main axis, was digitally established 100 m landward of the modern mean high tide line using the shoreline trend line. Thirty-five analysis transects with 30 m equal spacing were then perpendicularly generated along the baseline using the DSAS module’s automatic transect function (as shown in Figure 3). To ensure that the transects are perpendicular to the baseline, a systematic manual correction was carried out after the transects were automatically generated. The Angle Measurement Tool in ArcGIS was used to assess whether the angular relationships between each transect and its corresponding shoreline were normal. Adjustments were made via manual reorientation in DSAS where necessary to ensure all transects met the perpendicularity criterion [33,34]. Post-correction angle measurements confirmed compliance with the perpendicularity criterion. Each sample strip was configured to be 250 m long and smoothed through a 1 m coastline position to effectively capture the micro-terrain fluctuations of the beach [35].

WLR, EPR, and NSM were calculated for shoreline positional changes using the DSAS core algorithm library on a per-sample basis. In DSAS analysis, the rate of change is measured in m/yr, and shoreline movement is measured in m, both of which are expressed as positive and negative values. Negative values indicate shoreline retreat, while positive values indicate shoreline progradation [36].

(1)

WLR

WLR determines the trend of shoreline displacement over time by linear regression analysis of coastline locations at various temporal intervals. However, in the calculation, the reliability or weights of different measurement points were taken into account to provide a more accurate estimate of the shoreline movement rate. This research calculates and analyzes the rate of change of several coasts over an extended duration. The formula of WLR is as follows (Equation (2)):

$$WLR=1{\left(e\right)}^2$$

(2)

In this equation, e denotes the shoreline uncertainty value.

(2)

EPR

EPR measures the rate of shoreline change between two positions (i, j), computed as the Net Shoreline Movement (NSM) divided by the elapsed time. The formula of EPR is as follows (Equation (3)):

$$E_{ij}={\displaystyle \frac{d_j-d_i}{△Y_{ij}}}$$

(3)

$E_{i,j}$ denotes the rate of change in the shoreline endpoint between consecutive years along a tangent line extending from the baseline. $d_j$ represents the distance from the shoreline along the tangent line H to the baseline in period j. $d_i$ signifies the distance from the shoreline along the tangent line H to the baseline in period i. $\Delta Y_{i,j}$ indicates the difference in the number of years between the shorelines in periods j and i.

(3)

NSM

The net displacement of the coastline at each sample transect junction between the oldest and youngest shoreline locations. The formula of NSM is as follows (Equation (4)):

$$NSM=d_{young}-d_{old}$$

(4)

where $d_{young}$ refers to the distance from the baseline to the most recent coastline. $d_{old}$ refers to the distance from the baseline to the oldest coastline.

3.2.3. Method for Analyzing Sediment Particle Size

The technical protocol for sediment particle size analysis commenced with standardized field sampling procedures. In the laboratory, raw samples are treated using established techniques, and 10–20 g of typical material are measured. Organic debris was eliminated with a 30% hydrogen peroxide solution, while carbonate minerals were dissolved with a 10% dilute hydrochloric acid solution. Following desalination using deionized water, a 0.5% solution of sodium polyphosphate was introduced. The sample was subjected to 20 kHz ultrasonic agitation for 15 min to achieve complete particle dispersion. Particle size characteristics were assessed by laser diffraction, image analysis and sieving, in compliance with Marine Survey Standards (GB/T 12763.8-2007) [37]. During the method, quality control measures (including reagent blanks, duplicate analyses, and calibration with reference materials) were employed to guarantee data dependability, thereby facilitating correct interpretation of the sedimentary environment. The particle size distribution of tiny particles was assessed using a laser particle size analyzer. The device employed was a Beckman Coulter LS13 320 laser particle size analyzer produced by Malvern Instruments (UK) in the United Kingdom. The device quantifies particle dimensions from 0.4 to 2000 $\mathsf{\mu }$m, exhibiting a resolution of 0.01 $\Phi$ and a relative error of less than 5% for repeated assessments. The particle size distribution of coarse particles was assessed using sieve analysis. Particle size analysis focused on five key parameters, namely median particle size (Md$\phi$), average particle size (Mz), sorting coefficient ($\sigma$), skewness (Sk), and kurtosis (Kg) [38]. The sorting coefficient ($\sigma$) evaluates sediment sorting, with $\sigma$ < 0.35 denoting good sorting and $\sigma$ > 0.50 indicating poor sorting. Skewness (Sk) describes the symmetry of sediment distribution and kurtosis (Kg) indicates the sharpness of sediment particle distribution [39]. Grain size analysis was performed utilizing the Folk & Ward (1957) technique for parameter computation and the correlation between sediment properties and shoreline dynamic evolution was then examined based on the computed findings [39].

3.3. Error Analysis and Uncertainty Assessment

Methodological uncertainties were systematically evaluated to ensure robustness. Landsat ($\pm 10$ m) and Sentinel-2 ($\pm 5$ m) shoreline positional errors were reduced to $\pm 5$ m using CoastSat’s sub-pixel extraction, while tidal corrections (MIKE model; $\pm 0.1$ m error) standardized vertical positions. Thirty-five shore-perpendicular transects (30 m spacing, <5^∘ deviation) and a validated baseline (100 m landward of MHWL; <2 m uncertainty) ensured consistent shoreline change measurements. Sediment analytical errors from laser diffraction (<5% error for $0.4$–2000 $\mathsf{\mu }$m) and sieving ($\pm 2\%$ for >2000 $\mathsf{\mu }$m) were quantified. Despite the existence of technical and equipment limitations leading to unavoidable errors, these errors do not affect the statistical significance of the final results regarding the extreme shoreline changes.

4. Results and Discussion

4.1. Characterization of Beach Shoreline Changes at Long-Term Scales

From 2000 to 2010, the temporal evolution of the shoreline demonstrates significant spatial heterogeneity, specifically along the northern, central, and southern subsegments of the studied coastal area And distinct stage-specific transformation patterns across the study period (Figure 4).

The evolutionary process of the shoreline in the study area falls into three phases, closely linked to human activities. From 2000 to 2010, shoreline changes were minor, presumably because the northern reclamation projects were still unfinished and their impact remained limited. After 2010, rapid shoreline changes may have been triggered by altered local hydrodynamics, possibly due to the completion of reclamation. Ecological restoration projects implemented in 2015 might have played a role in slowing shoreline changes, with their effects seemingly becoming evident after 2020.

By integrating dynamic landform classification with clustering analysis of shoreline change rates, the 25-year shoreline evolution in the study area can be delineated into three functional geomorphic units, reflecting varying dominant processes. They include sediment-dominated zones characterized by shoreline retreat, transition zones undergoing dynamic adjustment, and shoreline retreat-prone zones exhibiting persistent shoreline retreat. Shoreline movement with a rate greater than 0.05 m/y is defined as accretionary, while a rate less than −0.05 m/y is classified as erosive, and the range between −0.05 and 0.05 m/y is considered stable.

Based on the spatial distribution of the three identified shoreline change units, the study area is subdivided into three regions using sample transect No. 19 as the boundary between landform types. The northern region comprises sample Transects 0–18, and the central region includes Transects 19–20. The southern region covers Transects 21–35.

As shown in Figure 5, the values of WLR, EPR, and NSM for the northern region (sample Transects 0–18) ranged from 0.05 to 3.29 m/yr, 0.05 to 0.93 m/yr, and 0.35 to 74.76 m. All three indicators exhibited positive values, indicating a consistent seaward advancement of the coastline over the past 25 years. This pattern reflected a predominantly depositional environment in the northern sector, corresponding to the sediment-dominated landform unit identified in the spatial classification. Combined analysis of the NSM data (Figure 5c) and historical imagery (Figure 4) indicated the presence of commonly called beach cusps in sample Transects 8 and 11. Transect 8 exhibits particularly pronounced sediment shoreline accretion, marked by a maximum value of NSM of 74.76 m. Therefore, this study defined the northern 0–18 sample belt area as the shoreline progradation-dominated tend zone (sediment-dominated zones characterized by shoreline accretion). This classification is supported by the consistently positive WLR, EPR, and NSM values, as well as the observed cape-like shoreline protrusions, which are indicative of wave refraction in front of the sector generating this morphology. According to the three standard evolutionary stages identified in the research domain, it is deduced that the alterations in the northern 0–18 sample transects display a two-stage pattern. The initial phase (2000–2010) was primarily influenced by natural dynamics, whereas the subsequent phase (post-2010) experienced heightened human intervention. This has led to an increase in the rate of shoreline. However, post-2020, the shoreline progradation rate diminished, indicating the influence of natural forces (such as wave action, tidal currents) in reestablishing equilibrium in the artificially modified landscape due to limitations on sediment supply.

In the central transects (19–20), absolute values of both WLR and EPR consistently remain below 0.5 m/yr, while NSM fluctuates between minor accretion (0.35 m) and a small shoreline retraction (−0.01 m) (Figure 5). These minimal changes align with the inherent dynamic equilibrium of natural beaches, where shorelines undergo subtle fluctuations rather than maintaining static positions. Such variability arises from natural drivers including wave-tide interactions, seasonal energy shifts, and local sediment transport. The observed NSM range directly reflects this expected background variability. This study designated the middle 19–20 sample transects region as a dynamic equilibrium tend zone (transition zones undergoing dynamic adjustment). The relative stability of this region is primarily maintained by the geomorphic structure of the regulating effect of the rotating tidal wave system. These features collectively serve as a natural buffer zone, modulating the north–south shoreline progradation–shoreline–shoreline retreat pattern driven by the interaction between coastal topography and regional hydrodynamic forces.

The WLR, EPR, and NSM values for the southern 21–35 sample regions vary from −0.95 to −0.23 m/yr, −0.35 to −0.05 m/yr, and −0.01 m to −27.41 m. Figure 5 illustrates that the WLR, EPR, and NSM values for the southern 21–35 sample regions were all negative, signifying that the shoreline in this region has retreated inland during the past 25 years. According to the NSM data presented in Figure 5c and the historical photographs in Figure 4, the largest land retreat measured 27.41 m, situated near the open bay of sample transects No. 32. This study identified sample transects No. 21–35 in the southern region as places dominated by shoreline retreat (shoreline retreat-prone tend zones) based on the aforementioned analytical results. Based on the three standard evolutionary phases identified within the study area, the shoreline retreat process can be characterized as occurring in two distinct stages. During the initial phase (2000–2010), shoreline retreat was minimal, and the coastline maintained relative dynamic stability. This stability is attributed to the combined effects of intensified winter wind and wave activity, which modulated sediment redistribution and intermittent sand mining operations that temporarily altered sediment availability. In contrast, the subsequent phase (post-2010) experienced a marked increase in shoreline retreat rates. This acceleration is primarily linked to anthropogenic modifications, notably the construction of port breakwaters and associated land reclamation projects, which disrupted natural sediment transport pathways. These interventions resulted in the concentration of wave energy within the open bay, thereby exacerbating shoreline retreat processes along the shoreline [40,41].

Figure 5 illustrates that the EPR and WLR of sample Transects 0–18 in the northern section of the study area (EPR = 3.2 m/yr vs. WLR = 5.6 m/yr) exhibit significant disparities, signifying a sudden alteration in the sedimentary dynamics of this region. This anomaly may be associated with the reconfiguration of the sedimentary environment induced by the land reclamation project in 2009. Spatially, the high shoreline progradation rate in sample transect No. 5 in the north may be attributed to the interception effect of wave energy dispersion at the cape on sediments. Meanwhile, the strong shoreline retreat in the open bay of sample transect No. 32 in the south is probably related to the insufficient total amount of sediment transported along the coast due to artificial structures. In terms of time, after 2010, the intensity of human activity interference with the natural sedimentary dynamic system significantly increased, becoming the core driving force behind the strengthening of the north–south differentiation pattern. According to the analysis results, the shoreline evolution in the study area shows a significant spatial differentiation pattern of northern siltation, southern shoreline retreat, and central buffering. This may suggest that the littoral system is heterogeneously driven by internal and external forces, potentially resulting in relatively independent evolutionary patterns. Such spatial diversity could be closely linked to variations in sediment supply, hydrodynamic conditions, and anthropogenic interventions across coastal segments—a conclusion further supported by Liu Changfeng’s wave numerical simulation study in the Lvshun New Port tourism leisure area [23].

4.2. Characteristics of Beach Shoreline Evolution Under Human Influence

4.2.1. The Impact of Land Reclamation Projects on the Evolution of Beach Coastlines

Land reclamation projects are an important means of human intervention in coastal systems, and the changes in coastal dynamics they cause have a significant and lasting regulatory effect on shoreline evolution. This paper was based on the interpretation of multi-source remote sensing data and quantitative analysis of coastline evolution parameters (WLR, EPR, NSM) from 2010 to 2024. Subsequently, it combined the 1985–2010 historical Landsat image time-space sequence to reveal the differentiated response mechanisms of coastline evolution and land reclamation projects in the study area.

This work examined the influence of the temporal and geographical development of land reclamation projects on abrupt alterations in shoreline morphology by analyzing historical photos spanning many years before and following the projects. According to the image in Figure 6, the land reclamation project in the northern part of the study area began with cofferdam construction in 2008 and was completed in 2010, forming an artificial cape with a large area of reclaimed land. Image comparison indicates that before land reclamation (2005), the northern shoreline exhibited a regular arc. But after land reclamation, the artificial cape extended approximately 1.3 km into the sea, altering the local wave energy distribution pattern. Figure 7c demonstrates that the NSM values in the northern 0–18 sample belt area range from 0.35 to 74.16 m. The maximum offshore advance is 74.16 m (sample belt No. 8), with an average annual offshore advance of 4.95 m. This region is located on the east side of the reclaimed land. Due to the wave shadow effect, a sedimentary advantage zone has formed, causing the coastline to advance seaward over the past 15 years. In the center 19–21 sample transect region, the NSM values vary from 0.35 to 15 m, reflecting the combined influences of natural evolution and anthropogenic factors. In the southern 21–35 sample transects, the NSM values for sample Transects 21–35 in the southern region range from −0.48 to −27.14 m. These data indicate that the shoreline has been in a state of retreat over the past 15 years.

This section analyzes nonlinear differences in shoreline evolution parameters and their potential response to sediment supply pattern changes. Figure 7 reveals that WLR values in the northern shoreline progradation tend zone (0.05–3.29 m/yr) consistently show positive accretion, though the range (3.24 m/yr) appears significantly greater than EPR values (0.88 m/yr). Furthermore, the disparity between WLR and EPR rates suggests exponential escalation with land reclamation advancement. These observations may indicate a shift from natural continuous sediment replenishment to engineering-induced pulse-type transport post-construction of the artificial cape. The amalgamation of historical shorelines (Figure 4 and Figure 6) reveals rapid coastal changes during 2010–2020 land reclamation, showing correspondence with WLR trends. This potentially supports anthropogenic impacts on accelerated shoreline progradation through source material modifications. Collectively, these findings suggest land reclamation projects may not only alter coastline morphology but could also trigger abrupt sediment supply changes by potentially interrupting natural shoreline progradation pathways [42,43].

This section examines morphological evolution and structural characteristics in the southern shoreline retreat tend zone, aiming to reveal spatial variability in retreat responses along the southern coastline, with particular interest in potential indirect influences from anthropogenic engineering. Historical imagery (Figure 6) suggests this zone was relatively stable pre-project, though post-2010 coastal currents appear deflected by the northern artificial cape. Notably, the artificial groyne at sample Transect No. 32 seems to have gradually shrunk and disintegrated near the artificial viewing platform, coinciding with what may be a negative coastal sediment transport balance. This pattern shows correspondence with the spatial distribution of WLR values (−1.79 to −0.23 m/yr) and maximum land retreat (27.14 m) in Figure 7. These observations imply that land reclamation could indirectly exacerbate southern shoreline retreat by potentially disrupting regional sediment budgets. Collectively, findings across northern advancement tend zones, central dynamic, and southern retreat tend zones suggest land reclamation may trigger dynamic reorganisation with cross-zone transmission effects [44,45].

This discontinuous transformation characteristic starkly contrasts with the long-term relative stability of the coastal stretch prior to the project, suggesting that the land reclamation initiative may have disrupted the original sediment transport equilibrium by altering the regional hydrodynamic pattern. Land reclamation structures created promontory-like topography that exerts a significant hydrodynamic shielding effect on coastal flow fields. This shielding generates a downstream hydrodynamic shadow zone characterized by sharp variations in flow velocity gradients. The energy dissipation within the flow field systematically reduced the transport capacity of both suspended and bedload sediments. This promoted sediment concentration within the shadow zone through density-driven deposition, ultimately forming characteristic artificial feather-shaped sedimentary formations. This mechanism arises from the gradient effect of coastal sediment transport [46,47,48].

Following the 2008 reclamation project, the study area experienced shoreline retreat in its southern segment and shoreline advancement in the northern segment. This spatial pattern parallels documented outcomes at Wolcheon Beach. There, large-scale reclamation for the Samcheok LNG terminal disrupted longshore sediment transport, resulting in near-complete sand beach depletion [44]. Consistent findings emerged from research on Jeddah’s coast, where land reclamation activities were linked to significant shoreline advancement adjacent to port areas [43].

Long-term shoreline evolution exhibited distinct morphological phases driven by extreme events, lacking seasonal cyclicity. This pattern aligns with the 50-year Bengello Beach study, which identified five discrete beach-foredune phases triggered by major storms, including a significant 1974 event [43]. Both studies confirm absent seasonal volumetric fluctuations, underscoring episodic high-energy disturbances as primary controls on sandy coastline evolution. Human activities further drive differential littoral responses, reflecting segment sensitivity to external forcing as key manifestations of morphological diversity.

4.2.2. Impacts of Ecological Restoration Projects on Beach Shoreline Evolution

As a typical human-intervened coastal system, the shoreline evolution process of World Peace Park Beach exhibits multi-stage composite driving characteristics. Since the end of the twentieth century, this beach has been subject to the combined effects of persistent changes in sediment supply and increased dynamic shoreline retreat due to abnormal sediment transport rates in the river basin and frequent extreme weather events. The tourism-oriented beach nourishment project implemented in 2005 involved the deposition of non-native fine-grained sediments to construct an artificial beach. However, due to the lack of consideration for the local wave–current coupled hydrodynamic conditions, significant sediment loss occurred in the nourished beach body (as detailed in Section 2). The project utilized its optimized hydraulic roughness characteristics to suppress the coastal current transect’s effect and the rate of shoreline retreat.

Figure A1 indicates that during the period from the completion of the land reclamation project in 2010 to the completion of the ecological restoration project in 2016, the maximum shoreline progradation volume in the northern shoreline progradation area reached 47.76 m (sample Transect 8). The annual average shoreline progradation rate was 7.96 m/yr. During the same period, the maximum shoreline retreat in the shoreline retreat zone was −19.58 m (sample Transect 32), with an annual average shoreline retreat rate of −3.26 m/yr. From 2016 to 2024, the ecological restoration period saw a maximum shoreline progradation volume reduction to 26.20 m (a 45% decrease), with an average annual rate of 3.28 m/yr. The maximum shoreline retreat in the shoreline retreat zone was limited to −7.56 m (sample Transect 32). The annual average rate diminished to −0.95 m/yr. The data analysis indicates that during the research period, shoreline dynamics demonstrated a stable dynamic equilibrium. The coefficient of variation for shoreline retreat/shoreline progradation rates declined by 58–63% relative to pre-engineering values.

In 2015, the ecological restoration project promoted the synergistic interaction of porosity—energy dissipation—landform feedback through the phased implementation of pebble beach berms and permeable submerged dikes. The annual shoreline progradation rate in the northern zone diminished from 7.90 m/yr to 3.28 m/yr after restoration (2016–2024), whereas the shoreline retreat rate in the southern region decelerated from −3.26 m/yr to −0.95 m/yr (Figure 8). This coastal stabilization trend suggests the porosity—energy dissipation—landform feedback mechanism may fundamentally mediate dynamic–morphological interactions [49].

4.3. Characterization of Shoreline Evolution Under the Influence of Natural Disasters

Typhoons Haitang (Figure A2) and Noru (Figure A3) impacted the area with sustained winds of 28 m/s and wave heights > 4 m. Their tracks converged near the Bohai Strait, generating storm surges that elevated water levels by 1.5 m. Haitang formed near the Nansha Islands on July 28th, weakened after making landfall in Taiwan, and weakened into a residual circulation over Jiangxi on August 1st, thus being unnumbered. At this point, the remnants of Haitang continued to head north. At the same time, Typhoon Noru moved northwestward towards the southern waters of Japan, forming a closed high-pressure center on the ocean surface between the residual circulation of Noru and Haitang. On August 3rd during the day, as Typhoon Noru continued to move northwestward, the high-pressure center broke. At night on August 3rd, the peripheral residual circulation of Typhoon Haitang and Typhoon Noru began to interact [28]. This study delineated four fixed transects along the beach, determined by beach length and topography, to collect sediment samples before and after a typhoon. The objective was to examine the beach’s response to and resilience against storm surges, thereby facilitating the monitoring of coastline alterations after a storm surge. A comparative investigation of sediment characteristics before and after the typhoon indicates the impact of typhoons on beach shoreline progradation and shoreline evolution.

The median grain size ($\phi$) indicates the central tendency of the sediment distribution and reflects average transport energy. Coarse sediments (negative $\phi$, >1 mm) occur in high-energy environments (e.g., high tide zones), while fine sediments (positive $\phi$, <1 mm) occur in low-energy settings (e.g., low tide zones). The mean grain size ($\phi$) indicates depositional capability, where lower values denote coarser grains and higher values finer grains. The sorting coefficient measures particle size uniformity, with higher values indicating poorer sorting. Skewness denotes the comparative location of the mean and median values, illustrating the energy variation throughout the shoreline progradation process. Frequency curves of sediments generated in various sedimentary settings have distinct morphologies. The skewness coefficient of a normal distribution frequency curve equals zero. In this study, grain size parameters were calculated using the Folk–Ward graphical method. If the value is below zero, the mean is positioned to the right of the median. If it is exceeds zero, the mean deviates to the left of the median. The skewness symbol only reflects the tail direction, while the overall fineness of the sediment particle size is determined by the median. The peak shape indicates the sharpness of the particle size distribution and can provide insights into sediment sorting, variations in transport medium energy and the depositional environment. Kg > 1 indicates narrow distributions and effective sorting, suggesting stable environments. Kg < 1 indicates broad distributions and poor sorting, reflecting sediment mixing or dynamic disturbance [50,51,52,53].

Figure A4 indicates that the mean median particle size (Md) at high tide increased by 0.075 following the typhoon. The overall particle size remained above 2000 $\mathsf{\mu }$m. This indicates that the sediment particles in this region are somewhat refined. Nevertheless, the sediment type remained at the gravel level and the coarse-grained attributes did not undergo any significant alteration. The mean particle size (Mz) increased from −4.3075 to −4.285, a change of 0.0225. This indicates that the particle size gradually transitioned towards finer particles. The sorting coefficient ($\phi$i) rose from 0.405 to 0.425, an increment of 0.02. This indicates that the grain size composition of the sediments tended o be diversified, and the sorting performance was relatively reduced. The skewness coefficient (Sk) decreased from 0.1250 to 0.0375. This trend toward weaker positive skewness signifies a reduction in fine-grained tail components. The kurtosis (Kg) rose from 0.9025 to 0.9925, reflecting an increase of 0.09. The peak of the grain size distribution curve tends to be concentrated. It indicates a relative increase in the material content of a specific particle size grade in the sediment and a tendency towards a more concentrated particle size distribution. The findings in Figure A4 and

Table A3. Grain size structure of beach sediments before a typhoon (P01 and P02).

	Station	Grain Size Fractions	Gravel (G)	Sand (S)					Silt (T)				Clay (Y)
		Coarse Classification	Fine Gravel	Coarse Sand		Middle Sand	Fine Sand		Coarse Silt		Coarse Classification		Fine Gravel			Coarse Sand
		Fine Classification	Fine Gravel	Very Coarse Sand	Coarse Sand	Middle Sand	Fine Sand	Fine Classification	Fine Gravel	Very Coarse Sand	Coarse Sand	Middle Sand	Fine Sand		Fine Classification	Fine Gravel
		Abbreviations	FG	VCS	CS	MS	FS	VFS	CT	MT	FT	VFT	CY		FY	Primary and Secondary Components
		μm	>2000	2000∼1000	1000∼500	500∼250	250∼125	125∼63	63∼32	32∼16	16∼8	8∼4	4∼2	2∼1	<1	(Participation ≥20%)
		$\mathit{\phi }$	<−1	−1∼0	0∼1	1∼2	2∼3	3∼4	4∼5	5∼6	6∼7	7∼8	8∼9	9∼10	10∼11
P01	1 (High tide level)	Volume Ratio (%)	100.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	G
		Grain Size Classes (%)	100.00	0.00					0.00				0.00
	2 (Mid tide level)	Volume Ratio (%)	0.00	0.00	0.50	1.62	78.73	15.77	0.92	0.48	0.52	0.48	0.35	0.36	0.27	S
		Grain Size Classes (%)	0.00	96.62					2.40				0.98
	3 (Low tide level)	Volume Ratio (%)	0.00	0.00	0.00	0.25	83.48	10.63	0.96	0.80	1.24	1.12	0.66	0.44	0.42	S
		Grain Size Classes (%)	0.00	94.36					4.12				1.52
P02	4 (High tide level)	Volume Ratio (%)	100.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	G
		Grain Size Classes (%)	100.00	0.00					0.00				0.00
	5 (Mid tide level)	Volume Ratio (%)	100.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	G
		Grain Size Classes (%)	100.00	0.00					0.00				0.00
	6 (Low tide level)	Volume Ratio (%)	100.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	G
		Grain Size Classes (%)	100.00	0.00					0.00				0.00

,

Table A4. Grain size structure of beach sediments before a typhoon (P03 and P04).

	Station	Grain Size Fractions	Gravel (G)	Sand (S)					Silt (T)				Clay (Y)
		Coarse Classification	Fine Gravel	Coarse Sand		Middle Sand	Fine Sand		Coarse Silt		Coarse Classification		Fine Gravel			Coarse Sand
		Fine Classification	Fine Gravel	Very Coarse Sand	Coarse Sand	Middle Sand	Fine Sand	Fine Classification	Fine Gravel	Very Coarse Sand	Coarse Sand	Middle Sand	Fine Sand		Fine Classification	Fine Gravel
		Abbreviations	FG	VCS	CS	MS	FS	VFS	CT	MT	FT	VFT	CY		FY	Primary and Secondary Components
		μm	>2000	2000∼1000	1000∼500	500∼250	250∼125	125∼63	63∼32	32∼16	16∼8	8∼4	4∼2	2∼1	<1	(Participation ≥20%)
		$\mathit{\phi }$	<−1	−1∼0	0∼1	1∼2	2∼3	3∼4	4∼5	5∼6	6∼7	7∼8	8∼9	9∼10	10∼11
P03	7 (High tide level)	Volume Ratio (%)	100.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	G
		Grain Size Classes (%)	100.00	0.00					0.00				0.00
	8 (Mid tide level)	Volume Ratio (%)	86.25	0.89	5.49	5.46	1.50	0.14	0.06	0.06	0.04	0.04	0.03	0.02	0.02	G
		Grain Size Classes (%)	86.25	13.48					0.20				0.07
P03	9 (Low tide level)	Volume Ratio (%)	100.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	G
		Grain Size Classes (%)	100.00	0.00					0.00				0.00
P04	10 (High tide level)	Volume Ratio (%)	100.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	G
		Grain Size Classes (%)	100.00	0.00					0.00				0.00
	11 (Mid tide level)	Volume Ratio (%)	98.26	1.21	0.36	0.05	0.04	0.02	0.01	0.02	0.01	0.01	0.01	0.00	0.00	G
		Grain Size Classes (%)	98.26	1.68					0.05				0.01
	12 (Low tide level)	Volume Ratio (%)	81.67	3.21	5.16	3.26	3.63	1.34	0.62	0.42	0.27	0.19	0.12	0.06	0.05	G
		Grain Size Classes (%)	81.67	16.60					1.50				0.23

,

Table A5. Grain size structure of beach sediments after a typhoon (P01 and P02).

	Station	Grain Size Fractions	Gravel (G)	Sand (S)					Silt (T)				Clay (Y)
		Coarse Classification	Fine Gravel	Coarse Sand		Middle Sand	Fine Sand		Coarse Silt		Coarse Classification		Fine Gravel			Coarse Sand
		Fine Classification	Fine Gravel	Very Coarse Sand	Coarse Sand	Middle Sand	Fine Sand	Fine Classification	Fine Gravel	Very Coarse Sand	Coarse Sand	Middle Sand	Fine Sand		Fine Classification	Fine Gravel
		Abbreviations	FG	VCS	CS	MS	FS	VFS	CT	MT	FT	VFT	CY		FY	Primary and Secondary Components
		μm	>2000	2000∼1000	1000∼500	500∼250	250∼125	125∼63	63∼32	32∼16	16∼8	8∼4	4∼2	2∼1	<1	(Participation ≥20%)
		$\mathit{\phi }$	<−1	−1∼0	0∼1	1∼2	2∼3	3∼4	4∼5	5∼6	6∼7	7∼8	8∼9	9∼10	10∼11
P01	1 (High tide level)	Volume Ratio (%)	100.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	G
		Grain Size Classes (%)	100.00	0.00					0.00				0.00
	2 (Mid tide level)	Volume Ratio (%)	0.00	0.00	0.00	10.13	78.98	15.75	1.20	1.12	0.99	0.80	0.53	0.42	0.30	S
		Grain Size Classes (%)	0.00	94.64					4.11				1.25
	3 (Low tide level)	Volume Ratio (%)	0.00	0.00	0.00	0.13	87.54	17.76	1.06	0.39	0.24	0.18	0.19	0.30	0.22	S
		Grain Size Classes (%)	0.00	97.42					1.87				0.71
P02	4 (High tide level)	Volume Ratio (%)	100.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	G
		Grain Size Classes (%)	100.00	0.00					0.00				0.00
	5 (Mid tide level)	Volume Ratio (%)	78.46	2.59	2.99	2.83	11.93	0.94	0.11	0.04	0.02	0.01	0.02	0.05	0.01	G
		Grain Size Classes (%)	78.46	21.28					0.18				0.08
	6 (Low tide level)	Volume Ratio (%)	97.47	1.75	0.36	0.04	0.24	0.10	0.01	0.01	0.01	0.01	0.00	0.00	0.00	G
		Grain Size Classes (%)	97.47	2.49					0.04				0.00

and

Table A6. Grain size structure of beach sediments after a typhoon (P03 and P04).

	Station	Grain Size Fractions	Gravel (G)	Sand (S)					Silt (T)				Clay (Y)
		Coarse Classification	Fine Gravel	Coarse Sand		Middle Sand	Fine Sand		Coarse Silt		Coarse Classification		Fine Gravel			Coarse Sand
		Fine Classification	Fine Gravel	Very Coarse Sand	Coarse Sand	Middle Sand	Fine Sand	Fine Classification	Fine Gravel	Very Coarse Sand	Coarse Sand	Middle Sand	Fine Sand		Fine Classification	Fine Gravel
		Abbreviations	FG	VCS	CS	MS	FS	VFS	CT	MT	FT	VFT	CY		FY	Primary and Secondary Components
		μm	>2000	2000∼1000	1000∼500	500∼250	250∼125	125∼63	63∼32	32∼16	16∼8	8∼4	4∼2	2∼1	<1	(Participation ≥20%)
		$\mathit{\phi }$	<−1	−1∼0	0∼1	1∼2	2∼3	3∼4	4∼5	5∼6	6∼7	7∼8	8∼9	9∼10	10∼11
P03	7 (High tide level)	Volume Ratio (%)	91.53	1.33	1.68	2.29	2.85	0.22	0.05	0.01	0.01	0.01	0.01	0.01	0.01	G
		Grain Size Classes (%)	91.53	8.36					0.08				0.03
	8 (Mid tide level)	Volume Ratio (%)	91.53	1.33	1.68	2.29	2.85	0.22	0.05	0.01	0.01	0.01	0.01	0.01	0.01	G
		Grain Size Classes (%)	91.53	8.36					0.08				0.03
	9 (Low tide level)	Volume Ratio (%)	80.22	1.34	1.95	4.44	9.95	1.43	0.26	0.12	0.08	0.07	0.05	0.05	0.04	G
		Grain Size Classes (%)	80.22	19.11					0.53				0.14
P04	10 (High tide level)	Volume Ratio (%)	100.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	G
		Grain Size Classes (%)	100.00	0.00					0.00				0.00
	11 (Mid tide level)	Volume Ratio (%)	65.16	13.68	14.78	5.22	1.09	0.07	0.00	0.00	0.00	0.00	0.00	0.00	0.00	SG
		Grain Size Classes (%)	65.16	34.84					0.00				0.00
	12 (Low tide level)	Volume Ratio (%)	82.49	2.52	3.95	2.64	7.07	0.87	0.10	0.09	0.08	0.07	0.05	0.05	0.01	G
		Grain Size Classes (%)	82.49	17.06					0.34				0.11

demonstrate that the sediment type remains consistently gravel (G) both prior to and following the typhoon during high tide. The grain size statistical parameters show slight variation while remaining relatively stable. This indicates that sediment reaction amplitude in the high tide area was minimally restricted during harsh weather events.

Figure 9 shows that the mean median particle size (Md) at mid-tide increased by 0.85 following the typhoon, remaining within the coarse-grained range (>20,000 $\mathsf{\mu }$m), though the trend toward finer particle size became more pronounced. The mean particle size (Mz) increased significantly from −1.9825 to −1.035, a change of 0.9475. This indicates substantial refinement of the sediment and a shift in the grain-size structure toward fine-dominated compositions. The sorting coefficient ($\sigma$i) rose from 0.825 to 1.5075, an increase of 0.6825. This reflects greater complexity in grain-size composition and a marked decline in sediment sorting performance. The slight decrease in Sk from 0.395 to 0.37 signifies weakening fine-skewedness, implying reduced influence of fine particle tails and potential hydrodynamic sorting enhancement. The kurtosis (Kg) declined from 1.2125 to 1.2025, a change of 0.01. This indicates a shift toward a more platykurtic (flatter) distribution curve, with slightly reduced concentration and uniformity in particle size composition. As illustrated in Figure 10 and Table A3, Table A4, Table A5 and Table A6, changes in sediment type at the sampling sites further support the above findings. Among the four monitoring stations, two sites (Stations 5 and 11) experienced a transition in sediment type from gravel (G) to sandy gravel (SG) following the typhoon, which indicates a boundary shift in sediment classification. This was characterized by a decline in gravel content and a notable increase in the proportion of sand components, signaling a fundamental change in material composition. The sediment types at Stations 2 and 8 remained unchanged. But there were observable shifts in the proportions of silt (T) and sand (S), with a clear trend of increased fine material content. Given that cobble replenishment has been carried out in this area, which would limit the input of fine materials from land, it is presumed that the increased fine materials are likely of marine origin. From the perspective of parameter variation, the grain size response of sediments in the mid-tide zone under typhoon disturbance was significantly more pronounced than that in the high-tide zone. This suggests that the mid-tide area is situated within a zone of stronger hydrodynamics and wave action. Its sedimentary structure and material composition are prone to rapid adjustments owing to the influence of processes such as typhoon storm surges, wave upsurge, and backflow. Sediment refinement, sorting deterioration, and grain size distribution patterns tend to become more complex.

As shown in Figure 9, the mean median particle size (Md) in the low-tide zone increased by 0.88 following the typhoon. Although the overall grain size remained within the coarse-grained category, the trend toward particle refinement was significant. The mean particle size (Mz) rose from −2.340 to −1.1825 (an increase of 1.1575), representing the largest change among the three tidal levels. This indicates a clear fining of the sediment grain size after the typhoon. The sorting coefficient ($\sigma$i) increased from 0.89 to 1.375 (a rise of 0.485). This shift marked a transition from a relatively well-sorted sediment structure to a more complex and poorly sorted one, with a more dispersed grain-size distribution. The rise in Sk from 0.3175 to 0.4425 signifies intensified fine-skewedness, denoting increased fine particle accumulation in distribution tails. The kurtosis (Kg) decreased from 1.195 to 1.1025, a change of 0.0925. This indicates reduced concentration in grain-size distribution, a flatter distribution curve and a more diversified grain-size composition. Data from Figure 11 and Table A3, Table A4, Table A5 and Table A6 further support these findings. At Stations 3, 6, 9, and 12, a general decrease in gravel (G) content and a noticeable increase in sand (S) components were observed. In particular, the gravel content at Station 9 dropped significantly from 100% to 80.22%, which indicates a marked loss of coarse material and an enrichment of fine particles. This transition toward finer sediment types further confirms that the low-tide zone exhibited the strongest response to typhoon disturbance. Being situated in the subtidal zone, the low-tide area is directly influenced by the wave base. During typhoon-induced storm surges, wave energy is intensely released, resulting in strong disturbance, resuspension, and redistribution of seabed sediments. Coarse-grained particles are frequently transported to the upper regions, while fine particles tend to remain and settle during low tide. This process initiates a systematic reconstruction of the granularity structure. The findings indicate that the intense hydrodynamic processes induced by the typhoon resulted in the most pronounced sedimentary response in the low tide zone. This response is characterized by significant grain-size fining, reduced sorting performance, enhanced skewness, and a complex grain-size distribution. Consequently, the low-tide zone emerges as the most sensitive and variable area within the beach system when exposed to extreme weather conditions.

In summary, the 2017 double typhoon event triggered significant sedimentary adjustments across tidal zones, directly driving short-term shoreline fluctuations. Post-typhoon sediment grain size analysis reveals a clear linkage between sedimentary structural changes and shoreline dynamics. Typhoon events have had a significant impact on the grain size characteristics of beach sediments in different tide zones. And this has promoted the reshaping of the spatial form and structure of the beach shoreline. In the high-tide zone, changes in grain size parameters were relatively minor. It was mainly reflected in the slight fining of the median particle size (Md) and mean particle size (Mz), increased kurtosis, and decreased skewness. The observed suggests that, despite the high-tide zone being susceptible to storm surge uprush, sediment disturbance remains minimal. Consequently, the structural integrity of the shoreline was largely preserved, with no significant alterations to the overall shoreline position. In contrast, the mid-tide zone experienced markedly intensified impacts from typhoon-induced disturbances. It showed that the particle size composition was significantly refined and both the mean particle size (Mz) and the sorting ($\sigma$i) capacity were observed. And sediment type replacement occurred at some stations. This structural change reflected the improved plasticity of the shoreline morphology in the mid-tidal zone. During typhoons, local shoreline retreat or siltation was more likely to occur, resulting in oscillations of the shoreline. Such responses often result in microscale shoreline oscillations or seaward advancement. In contrast, the low-tide zone functioned as the principal scouring region due to the synergistic effects of storm surge and vigorous wave activity. This zone demonstrated the most significant alterations in grain size parameters. Md and Mz are significantly elevated, ($\sigma$i) is elevated, Kg is decreased, and skewness is enhanced. Moreover, the proportion of gravel components markedly decreased at several stations, while the proportion of sand components significantly increased. This shift indicates a migration of coarse material and an enrichment of fine material, resulting in substantial adjustments within the sedimentary system. These alterations rapidly induce instability in the coastal substrate, leading to comprehensive modifications in the morphology of beach transects, such as shoreline retreat and morphological reshaping [54].

Synergistic parameter variations suggest that typhoon-induced hydraulic winnowing may drive sediment deficit, exacerbating foreshore retreat. Localized shoreline progradation could be generated through littoral hydrodynamic segregation (combining wave scouring and backwash deposition). The coupled landward advection of coarse clasts and flocculative retention of fines might trigger ephemeral shoreline retrogression, positional oscillation, and bathymetric reconfiguration. This metastable sediment flux equilibrium may underscore the autogenic reorganization capacity of coastal morphodynamics under extreme climatic forcing [55].

Typhoon events triggered rapid shoreline retreat in the study area, followed by post-disturbance recovery. This resilience parallels coastal responses documented in central-southern Chile after the 2010 earthquake and tsunami [56,57]. There, despite severe initial perturbations, the shoreline recovered rapidly and maintained long-term stability. Both cases illustrate the high adaptive capacity of sandy coastal systems to extreme natural disturbances. This tidal zone-dependent response pattern indicates that identical natural disturbances may induce distinct morphodynamic adjustments across different littoral system components. Consequently, the spatiotemporal heterogeneity inherent in coastal morphological evolution is amplified.

5. Conclusions

This study utilized remote sensing imagery (2000–2024) and DSAS-calculated WLR, EPR, and NSM metrics to analyze coastal evolution. Quantitative multi-spatiotemporal analysis elucidated anthropogenic and natural disturbance mechanisms on beach dynamics. Pre/post-typhoon sediment comparisons demonstrated rapid shoreline retreat followed by recovery, highlighting intrinsic resilience. These findings provide science-based foundations for sustainable coastal management strategies in anthropogenically pressured sandy systems. The conclusions are as outlined below.

(1) This study identifies significant spatiotemporal patterns in shoreline evolution: spatially, distinct zones include a northern shoreline advancement-tended zone, a central transitional zone, and a southern shoreline retreat-tended zone; temporally, periods comprise a dynamic equilibrium phase (2000–2010), a high-intensity change phase (2010–2020), and an emergent dynamic equilibrium phase (2020–2024). Spatial differentiation appears strongly associated with variations in dominant regional processes: shoreline advancement in the north is likely linked to riverine sediment supply and land reclamation activities, while southern shoreline retreat is correlated with high wave-current energy, sea-level rise, and sediment deficit. The high-intensity change phase (2010–2020) coincides significantly with large-scale human interventions, notably land reclamation, suggesting a potential disruption of the prior equilibrium state. While the post-2020 emergent equilibrium may indicate an initial system adaptation to disturbance, its long-term stability remains uncertain under persistent natural and anthropogenic pressures. Collectively, these findings point to a substantial role of human activities in coastal evolution and highlight the need for spatially differentiated management strategies.

(2) Reclamation initiatives have modified the dynamic environment. Between 2008 and 2010, the artificial cape in the north induced shoreline progradation on the eastern side and shoreline retreat in the southern region, forming an engineering pulse sand replenishment and transportation mode. The ecological restoration project enhanced pore energy efficiency via pebble beach berms, decreased siltation in the north, alleviated shoreline retreat in the south, and stabilized the shoreline. Thus, the efficacy of structural interventions in harmonizing coastal dynamics is demonstrated through these observed changes.

(3) The analysis based on the particle size parameters of beach sediments in different tidal zones showed that strong storm surges and wave effects caused by typhoons significantly affect the particle size structure and composition characteristics of beach sediments. The particle size characteristics in the high-tide area change were limited. The sedimentary disturbance was relatively weak and the shoreline structure was relatively stable. In contrast, the mid-tide zone exhibited notable particle size refinement and diminished sorting capacity. The alterations in sediment type were indicative of the high plasticity of the shoreline morphology. The low-tide area was especially affected by the combined influence of waves and storm surges, with major variations in particle size parameters. The coarse-grained fraction decreased, whereas the fine-grained fraction increased. This sedimentary reorganization triggered basal instability along the coastal transects, thereby driving subsequent shoreline retreat and morphological realignment. Thus, the research results reveal that the sedimentary structural changes generated by typhoons are an important mechanism for the morphological evolution of beach shoreline and have a profound impact on the stability of the coastal zone.

The findings of this study, obtained from a 1 km semi-enclosed bay beach, hold broader relevance for global semi-enclosed sandy coast management. A key outcome is the identification of three distinct morphodynamic zones: a northern shoreline advancement zone, a central transitional adjustment zone, and a southern shoreline retreat-active zone. This spatial differentiation, primarily attributed to artificial headland configurations but potentially influenced by shoreline jetties, reflects established coastal responses to anthropogenic hydrodynamic disturbance. These zonal patterns yield sustainable management frameworks for small-scale tourist coasts globally. Furthermore, spatially differentiated responses across these zones during typhoon events were quantified, refining the mechanistic understanding of storm-driven sediment dynamics. Future work will specifically examine jetty impacts on localized hydrodynamics. Critically, cobble nourishment interventions were demonstrated to enhance coastal resilience. This methodology presents a scalable approach for reconciling tourism infrastructure requirements with long-term coastal sustainability in comparable mid-latitude monsoon environments.