Trends and Applications of Hydro-Morphological Modeling in Estuarine Systems: A Systematic Review of the Past 15 Years

Abstract

Estuaries are dynamic ecosystems with crucial environmental, economic, and social functions, driving extensive hydro-morphological research supported by numerical modeling. This study systematically reviews estuarine modeling applications over the past 15 years to identify commonly used tools, model configurations, and validation strategies, to examine regional trends in the application, and to explore and discuss the relative emphasis on hydrodynamic, sediment transport, and morphological modeling within the selected studies. Following the PRISMA 2020 methodology, a comprehensive search in Scopus and Web of Science identified 3926 articles, from which 197 met the eligibility criteria. Each study was analyzed to assess modeling software, mesh types, dimensional configurations, and validation parameters. Results indicate that DELFT3D is the most widely used tool, followed by TELEMAC and FVCOM, with a preference for two-dimensional models and structured meshes. Model accuracy, assessed through Skill Scores, confirms their reliability in representing estuarine dynamics. Additionally, findings reveal significant geographical disparities, with China leading research efforts, while Latin America and Africa remain underrepresented. This gap highlights the need to expand modeling efforts in these regions to enhance estuarine management and resilience. Strengthening numerical modeling in diverse contexts will improve the predictive capacity of hydro-morphological processes, supporting sustainable decision-making in estuarine environments.

1. Introduction

Estuaries are among the world’s most important natural systems due to their high environmental, social, and economic value, serving as hubs of biodiversity, natural barriers against coastal erosion, and sources of sustenance for local communities [1,2]. These environments are shaped by dynamic interactions between fluvial and coastal processes [3,4,5], which influence their morphology, hydrodynamics, and sediment dynamics [6]. River discharge dampens the tidal waves, enhances the ebb flow, and delivers sediment [7], while coastal influences, such as, littoral drift, storm surges, wind-driven waves, and sea-level-rise [8], pose significant disturbances to estuarine morphology and water quality [9].

The complexity of these systems and their extensive use has driven substantial scientific interest, transitioning from early conceptual models [10] to sophisticated laboratory experiments [11] and advanced numerical simulations [12]. In time numerical models have emerged as essential tools for analyzing estuarine hydrodynamics, with a wide range of software now available to model fluvial and coastal processes, sediment transport, and morphological changes [13,14]. This diversity of available tools underscores the equal importance of selecting the most suitable software to represent estuarine dynamics, configuring parameter values effectively to ensure model performance, and validating results to guarantee their reliability [15].

The application of precise numerical models offers significant benefits for estuarine management and research. For instance, studies have successfully predicted morphological changes in tidal inlets and sediment transport patterns, aiding the design of sustainable coastal protection structures [16]. In another case, numerical simulations in the Yangtze Estuary demonstrated the impacts of sea-level rise and sediment supply variations on deltaic morphology, providing critical insights for flood risk mitigation and navigation improvements [17]. Similarly, models have been applied in the Chesapeake Bay, enabling accurate assessments of storm surge dynamics and their influence on estuarine circulation [18]. These case studies illustrate how precise modeling results enhance decision-making processes, optimize resource management, and minimize environmental risks. Accurate simulations provide critical data for infrastructure planning, habitat conservation, and climate change adaptation, ensuring that estuarine systems remain resilient to both natural and anthropogenic pressures [19].

Comparing model outputs has proven to be a valuable approach to assessing the suitability and reliability of these tools [20,21,22]. This can be further supported by systematic analyses of existing studies, which identify common configurations and response ranges [23]. The objective of this study is to conduct a systematic review of numerical models applied to estuarine environments over the past 15 years, with a focus on identifying (i) the most commonly used software tools, (ii) the configurations employed (mesh type, dimensionality, morphological acceleration), and (iii) validation strategies applied in the literature. The review also seeks to identify regional trends in application and discuss the relative emphasis on hydrodynamic, sediment transport, and morphological modeling within the selected studies.

To achieve this, relevant literature was retrieved from scientific databases such as Scopus and Web of Science (WoS), providing a comprehensive overview of methodologies, applications, and outcomes in the field.

2. Materials and Methods

The systematic review was conducted using the PRISMA 2020 methodology [24] which is structured in three main stages to ensure a rigorous and comprehensive analysis of numerical models applied to estuarine systems: document collection (i.e., identification), processing based on defined criteria (i.e., screening), and analysis of selected documents (i.e., inclusion). These stages follow established systematic review protocols and provide a clear and organized framework for this study [25]. The corresponding verification checklist is provided in Table S1 of the Supplementary Material.

2.1. Document Collection—Identification

The first stage, identification, involved implementing a detailed search strategy, based on filters, to identify relevant research articles. The criteria applied in this study included a temporal scope, focusing on articles published in the last 15 years (2010–2024) to ensure the inclusion of recent advancements [26,27,28]. The scientific databases Scopus and Web of Science (WoS) were used for this process. These two platforms are widely recognized as leading citation databases, collecting and disseminating bibliometric data on research articles, journals, institutions, and individual authors [29]. Scopus and WoS are complementary tools; while Scopus provides broader journal coverage, WoS focuses on high-impact selective journals [30]. However, both databases are essential for systematic reviews, as they enable comprehensive searches and facilitate the efficient management of overlapping citations [31].

The selection of keywords used in the filtering process was directly guided by the objectives of the review, which focus on the application of numerical models to simulate morphodynamic evolution in estuarine systems. Given that morphodynamic modeling requires the representation of both hydrodynamic and sediment transport processes, the search strategy was designed to capture studies that explicitly addressed numerical modeling of estuarine morphodynamics, or studies that, while focused on hydrodynamics or sediment transport, were framed as steps toward or components of morphodynamic analysis. To this end, combinations of terms such as “estuar*”, “morphodynamics OR morphological changes OR sediment transport OR hydrodynamics”, and “numerical model*” were used to ensure thematic alignment with the scope of the study. While this approach may have excluded studies using alternative terminology, it was necessary to preserve the specificity and relevance of the resulting dataset.

Thus, the following filters were applied to both databases:

• WoS: TS = (estuar*) AND TS = (morphodynamics OR morphological changes OR sediment transport OR hydrodynamics) AND TS = (numerical model* OR model).
• Scopus: TITLE-ABS-KEY(estuar*) AND TITLE-ABS-KEY(morphodynamics OR morphological changes OR sediment transport OR hydrodynamics) AND TITLE-ABS-KEY(numerical model* OR model).

The names and DOIs of the articles were downloaded from both the WOS and SCOPUS databases to create a unified dataset. Duplicate articles present in both databases were identified and removed, ensuring that only one unique copy of each article was retained in the final dataset. The process of identifying and removing duplicates was performed by cross-checking repeated titles using an Excel spreadsheet.

2.2. Processing Based on Defined Criteria—Screening

The second stage is screening. Here the identified articles were screened and filtered for a second time using explicit eligibility criteria to ensure their quality and relevance. The thematic focus was limited to studies employing numerical modeling to analyze hydro-morphological processes in estuarine environments. To ensure comparability among studies, we excluded articles with simulation periods exceeding one year. This decision was based on the observation that long-term modeling efforts often address different research objectives—such as evaluating decadal-scale morphological trends, climate scenarios, or anthropogenic impacts—and tend to involve higher levels of uncertainty in boundary conditions and input data [32,33,34,35,36,37]. These long-term simulations differ substantially in structure, validation approach, and temporal resolution from short-term morphodynamic studies, which typically focus on tidal cycles, storm events, or seasonal patterns [38,39,40,41]. Limiting the review to studies with simulation periods of one year or less helped maintain thematic coherence and analytical consistency within the dataset. Only peer-reviewed journal articles were included in the analysis to maintain methodological rigor.

Table 1. Screening Criteria for Selecting Articles of Interest, with * indicating stages conducted manually and ** indicating stages completed automatically through search engines.

Language	English and Spanish
Document Type **	WoS: Articles, proceeding papers, and early access Scopus: Articles and conference papers.
Subject of Interest *	Maximum 1 year of modeling performed in estuaries.
Search Period **	From 2010 to 2024.
Access Type *	All articles meeting the criteria were selected from Open Access journals or accessible through a subscription held by the Universidad Católica de la Santísima Concepción.

summarizes the mentioned screening criteria.

This phase conducted a review focusing on the title and abstract of the articles within the dataset. This step aimed to exclude articles whose study areas were not focused on estuaries, as well as research in fields outside the scope of this study, such as chemistry and biology. However, articles where the title and abstract did not provide sufficient information were retained for further evaluation. Subsequently, a thorough analysis of the articles from the previous phase was conducted. Each article was read in its entirety, discarding those that did not align with the subject of interest.

2.3. Analysis of Selected Documents—Inclusion

The third and final stage focuses on the inclusion of articles that passed the screening phase and aligned with the proposed research goal. The selected documents were systematically analyzed by reading the full article to extract key information regarding the numerical models used. First, the information of interest to be extracted from each article was defined, corresponding to the title, author, publication year, estuary name and country, software used, mesh characteristics, topology, model validation parameters, and morphological acceleration factor. The information was organized in a tabular format using an Excel spreadsheet. The steps followed for the construction of the relevant information database were as follows:

(i) Storage of identification information (title, author, and publication year).

(ii) Analysis of the study area or methodology section, as applicable, to obtain the estuary name and country.

(iii) Thorough review of the methodology section to identify the model type, software used, mesh characteristics, topology, and morphological acceleration factor.

(iv) Review of the results section to extract statistical measures to evaluate the precision of the models, which indicates the correspondence between the model’s results and data obtained from field measurements. These were taken directly from the articles and were not explicitly calculated. This point of interest was addressed after analyzing the model configurations (iii) since it was only extracted from the most frequently used software.

Thematic classification of the reviewed studies was based on the modeling objectives and process integration described in each article. Hydrodynamic models (MH) were defined as those focusing solely on the simulation of flow characteristics (e.g., water level, velocity) without considering sediment dynamics. Sediment transport models (MT) included studies that modeled suspended or bedload transport but did not include morphological updating. Morphological models (MM) explicitly simulated temporal changes in bed elevation as a function of sediment transport, often using morphological updating routines or acceleration factors. In cases where multiple processes were included, studies were classified according to the highest level of physical integration present. This approach ensured consistency in thematic grouping and reflected the process hierarchy typical of morphodynamic modeling workflows.

Stages 1 through 3 are represented in Figure 1, which summarizes the provided description. The process concluded on 24 April 2024, therefore articles published after this date are not included in this study. The extracted information was summarized and will be presented in the results section.

3. Results

After applying the identification filters, 2645 articles were found in WoS and 2529 in Scopus. Duplicate articles were then removed (i.e., 1245), leaving an initial database of 3929 articles (WoS + Scopus). Figure 2 shows the publication trend indicating a sustained growth of publications for the initial database.

During the screening stage, a review of titles and abstracts based on the established criteria reduced the dataset to 686 articles. In the final inclusion stage, a comprehensive and detailed review further narrowed that dataset to 197 articles (Figure 1).

Appendix A provides a table summarizing the 197 articles in the final database, categorized based on eight aspects relevant to this research. These aspects were selected because they represent the traditional steps required to successfully complete a modeling process [42,43].

• Estuary,
• Estuary country,
• Modeling software,
• Model type (MH: Hydrodynamic only, MT: Sediment transport (Morphological without bathymetry updates), MM: Morphological with bathymetry updates),
• Mesh topology (structured or unstructured),
• Dimension (1D, 2D, or 3D),
• Morphological acceleration factor or hydrodynamic timestep (dt hydro) and morphological timestep (dt mor), as applicable,
• Validation parameter (Skill Score (SS)) used in the articles to validate the simulations, where: $S{S}_{vel}$ is the number of values calculated for velocity validation and $S{S}_{wl}$ is the number of values calculated for water level validation.

Geographically, the selected estuary studies are dominated by Asian countries (49%), followed by European (29%), North American (12%), Oceanic (5%), South American (4%), and finally African countries (2%), as shown in Figure 3.

The high number of studies conducted in Asia is due to research carried out in China (69% of all Asian articles), a country that has one of the world’s largest estuarine systems, the Yangtze River. This estuary is one of the most studied systems in the world [44], located on the eastern coast, and is one of the world’s longest rivers and the longest in China (6300 km) [45,46,47]. It also has one of the broadest estuaries, making it one of the most dynamic systems, both physically and biogeochemically [48,49].

In Europe, the dominance is less marked than in Asia; however, the United Kingdom leads in research, accounting for 25% of all European-identified articles. According to these results, the most studied estuary is the Severn Estuary, located in southwest England, noted for its potential for tidal energy generation due to having one of the largest tidal ranges in the world (a maximum of 14 m during spring tide at Avonmouth) [14,50,51,52].

In North America, research is predominantly conducted in the United States (91% of North American articles), where the most studied estuaries are the Columbia and the Hudson. The Columbia River, located in the northwestern United States, is the largest river system on the U.S. West Coast, with an annual discharge of ~7300 m³/s [16,53,54]. Meanwhile, the Hudson River, situated in the northeastern United States, has an estuary classified as partially mixed, with a tidal range varying from 0.5 to 2.0 m and discharge ranging from 200 to 2000 m³/s [55,56,57].

In Oceania, Australia leads with 78% of Oceanic articles, with the most studied areas being Currumbin and Shoalhaven. Currumbin, located in southeastern Queensland, is a wave-dominated stream with an annual discharge range from 6 to 70 Mm³/year [58,59,60]. The Shoalhaven River, in southeastern New South Wales, has a wave-dominated estuary, and its flow is regulated by the Tallowa Dam [61,62].

In South America, Brazil leads in research output, accounting for 63% of South American publications. The most studied system is the Santos estuarine system, located on the south-central coast of São Paulo state. This estuary is particularly significant as it hosts the largest port in Latin America [63,64].

Finally, in Africa, only two studies were found, focusing on the Bouregreg and Oum-Errabia estuaries. The Bouregreg River, in northwestern Morocco, spans 240 km with an average flow ranging from 3 to 84 m³/s and an average tidal range of 2.3 m [65,66]. The Oum-Errabia River, which flows near Azemmour, is the second-longest river in Morocco (550 km) with an average flow of 117 m³/s [12].

The majority of the investigated estuaries are large systems, reflecting their global significance as they support substantial human populations [67,68]. Additionally, estuaries with significant tidal ranges, which offer considerable potential for energy generation, as well as those with critical infrastructure for the local economy are notably emphasized. However, the results also reveal a notable underrepresentation of many highly relevant estuaries, as well as certain geographic regions, in the reviewed research.

Selected articles were grouped into three sections based on their themes: (i) Hydrodynamic Model (MH), (ii) Sediment Transport Model (MT), and (iii) Morphological Models (MM), which were further divided into: articles that presented models that declare a bathymetric acceleration factor (MM1) and those that did not (MM2). The classification indicated that 46.2% of the articles are MH, while 36.5% and 17.3% correspond to MT and MM, respectively (Figure 4).

The low number of models with bathymetry updates does not allow for statistical analysis of their respective morphological parameters. However, various studies indicate that effective morphological modeling is directly related to correct hydrodynamic configuration [69,70]. Consequently, analyzing hydrodynamic models enables the identification of a software’s suitability over others when conducting morphological analyses.

Figure 5 shows the types of models that are most frequently used in different countries. The countries with the highest number of MH articles are China (26.4%), the United States (15.4%), and Portugal (8.8%). Meanwhile, the highest number of MT articles comes from China (47.2%), the United States (9.8%), the United Kingdom (5.5%), and Vietnam (5.5%). For MM models, the most research is from China (23.5%), Taiwan (17.6%), Portugal (11.8%), and the United Kingdom (11.8%) (Figure 5).

3.1. Hydrodynamic Modeling Software (MH)

The analysis of each article’s methodology revealed a wide variety of hydrodynamic modeling software. The four most frequently used software tools, based on the number of times they were employed to model a specific study site for a given objective, are identified in Figure 6 and added up to 57% of the studies found. The remaining 43% of cases involve 36 software tools, each using less than 5% of the total number of studies, making their analysis statistically insignificant. These less commonly used software tools are detailed in

Table A1. Summary of Articles Systematically Searched.

Author	Country Classification	Estuary	Software	Model Type	Topology	Dimension	Morfac	Validation
[110]	Australia	Brisbane Estuary	Mike	MH	-	2D
[60]	Australia	Currumbin Creek	Delft3D	MT	Structured	2D
[111]	Australia	Currumbin Creek	Delft3D	MM	Structured	2D	MORFAC = 4
[112]	Australia	Fitzroy Estuary	FVCOM	MM	Unstructured	3D
[113]	Australia	Port Curtis Estuary	Delft3D	MT	Structured	3D
[62]	Australia	Shoalhaven Estuary	Delft3D	MH	Structured	2D
[61]	Australia	Shoalhaven Estuary	Delft3D	MH	Structured	2D
[114]	Bangladesh	Meghna estuary	Delft3D	MM	Structured	2D	MORFAC = 12
[115]	Bangladesh	Pussur River	HEC-RAS	MH	-	1D
[116]	Brazil	Amazonian Estuary	Delft3D	MH	Structured/Unstructured	1D/2D
[117]	Brazil	Caravelas Estuary	Delft3D	MH	Structured	2D
[118]	Brazil	Paranagua Estuary Complex	Delft3D	MT	Structured	3D
[63]	Brazil	Santos estuarine system	Delft3D	MH	Structured	3D
[64]	Brazil	Santos estuarine system	Delft3D	MH	Structured	2D/3D
[119]	Canada	St. Lawrence River Estuary	H2D2	MH	Unstructured	2D
[120]	Canada	St. Lawrence River Estuary	Non-commercial	MH	Unstructured	1D/2D
[121]	China	Yangtze estuary	FVCOM	MH	Unstructured	3D
[47]	China	Yangtze estuary	Delft3D	MH	Structured	2D
[122]	China	Yangtze estuary	FVCOM	MH	Unstructured	3D
[123]	China	Yangtze estuary	FVCOM	MT	Unstructured	3D
[124]	China	Huangmaohai Estuary	COAWST	MT	Structured	1D/3D
[125]	China	Huangmaohai Estuary	ECOM	MT	Structured	2D
[126]	China	Jiaojiang Estuary	Mike	MT	Unstructured	3D
[127]	China	Jiaojiang River Estuary	FVCOM	MT	Unstructured	3D		$S{S}_{vel}$ (20)
[13]	China	Lingdingyang Estuary	TELEMAC	MM	Unstructured	2D	dt mor = 10 s/dt hydro = 30 s	$S{S}_{wl} \left(12\right)$ $S{S}_{vel}$ (3)
[128]	China	Lingdingyang Estuary	TELEMAC	MT	Unstructured	2D		$S{S}_{wl} \left(11\right)$ $S{S}_{vel}$ (23)
[129]	China	Luanhe River Estuary	Non-commercial	MH	Unstructured	2D
[130]	China	Modaomen Estuary	TELEMAC	MH	Unstructured	2D
[131]	China	Modaomen estuary	TELEMAC	MM	Unstructured	2D	dt mor = 10 s/dt hydro = 10 s	$S{S}_{wl} \left(1\right)$ $S{S}_{vel}$ (1)
[132]	China	Oujiang Estuary	FVCOM	MT	Unstructured	3D		SSwl (1)
[133]	China	Oujiang Estuary	ADCIRC	MM	Unstructured	2D
[134]	China	Oujiang Estuary	EFDC	MT	Structured	3D
[135]	China	Pearl River Delta	TELEMAC	MM	Unstructured	2D		$S{S}_{wl} \left(1\right)$ $S{S}_{vel}$ (1)
[136]	China	Pearl River Delta	ROMS	MT	Structured	3D
[137]	China	Pearl River Delta	[138]	MH	-	1D
[139]	China	Pearl river Delta	Developed by the State Key Laboratory	MH	Unstructured	2D
[140]	China	Pearl River Delta	ECOM/Riv1D	MT	Structured	1D/3D
[141]	China	Pearl River Delta	TELEMAC	MT	Unstructured	2D
[142]	China	Pearl River Delta	SCHISM	MT	Unstructured	3D
[143]	China	Pearl River Delta	COAWST	MT	Structured	3D
[144]	China	Pearl River Delta	ECOM	MT	Structured	3D
[145]	China	Pearl River Delta	EFDC	MT	Structured	3D
[146]	China	Pearl River Delta	EFDC	MT	Structured	3D
[147]	China	Pearl River Delta	Delft3D	MM	Structured	2D	MORFAC = 1
[148]	China	Pearl River Delta	FVCOM	MH	Unstructured	2D
[149]	China	Pearl River Delta	Delft3D	MH	Structured	3D
[150]	China	Qiantang Estuary	Non-commercial	MH	Unstructured	2D
[151]	China	Qiantang Estuary	FVCOM	MT	Unstructured	3D
[152]	China	Qiantang Estuary	Non-commercial	MT	Unstructured	2D
[153]	China	Qiantang Estuary	FVCOM	MH	Unstructured	3D
[154]	China	Qiantang Estuary	FVCOM	MH	Unstructured	2D
[155]	China	Sheyang estuary	Delft3D	MT	Structured	2D
[156]	China	Xinyanggang River	Delft3D	MM	Structured	2D	MORFAC = 1
[157]	China	Yalu River Estuary	FVCOM	MM	Unstructured	3D
[158]	China	Yalu River Estuary	FVCOM	MT	Unstructured	3D
[159]	China	Yangtze estuary	TELEMAC	MH	Unstructured	2D
[160]	China	Yangtze Estuary	Non-commercial	MM	Unstructured	2D
[48]	China	Yangtze Estuary	Delft3D	MH	Structured	2D		$S{S}_{wl} \left(19\right)$ $S{S}_{vel}$ (20)
[161]	China	Yangtze Estuary	SWEM3D	MT	Unstructured	3D
[49]	China	Yangtze Estuary	Delft3D	MH	Structured	2D		$S{S}_{wl} \left(5\right)$ $S{S}_{vel}$ (6)
[17]	China	Yangtze Estuary	TELEMAC	MH	Unstructured	2D
[162]	China	Yangtze Estuary	Non-commercial	MH	Unstructured	2D
[163]	China	Yangtze Estuary	Delft3D	MT	Structured	2D
[164]	China	Yangtze Estuary	Delft3D	MH	Structured	2D
[165]	China	Yangtze Estuary	-	MH	Unstructured	2D
[166]	China	Yangtze Estuary	Non-commercial	MT	-	1D
[167]	China	Yangtze Estuary	Developed by the State Key Laboratory	MH	Unstructured	2D
[168]	China	Yangtze Estuary	Non-commercial	MT	Unstructured	2D
[169]	China	Yangtze Estuary	COAWST	MT	-	3D
[170]	China	Yangtze Estuary	Delft3D	MH	Unstructured	3D		$S{S}_{wl} \left(23\right)$ $S{S}_{vel}$ (12)
[171]	China	Yangtze Estuary	FVCOM	MH	Unstructured	2D
[46]	China	Yangtze Estuary	Delft3D	MT	Structured	3D
[172]	China	Yangtze Estuary	Mike	MH	Unstructured	2D
[45]	China	Yangtze Estuary	CCHE	MH	Structured	3D
[173]	China	Yellow River	ECOM	MT	Structured	3D
[174]	China	Yellow River	FVCOM	MT	Unstructured	3D
[175]	China	Yellow River	Delft3D	MT	Structured	3D		$S{S}_{vel}$ (1)
[176]	China	Yellow River	TELEMAC	MT	Unstructured	2D
[177]	China	Yellow River	FVCOM	MT	Unstructured	3D
[178]	China	Yellow River/Yangtze River/Mekong River	Delft3D	MT	Structured	2D
[179]	China	Zhujiang River Estuary	ROMS	MT	Structured	3D
[180]	China	Zhujiang River Estuary	ROMS	MT	-	3D
[181]	Colombia	Magdalena River	OpenFlow Flood	MT	Structured	3D
[182]	France	Authie Estuary	TELEMAC	MT	Unstructured	2D
[183]	France	Charente Estuary	MARS3D	MH	-	2D
[184]	France	Gironde Estuary	TELEMAC	MH	Unstructured	3D
[185]	France	Gironde Estuary	TELEMAC	MT	Unstructured	2D
[186]	France	Gironde estuary	TELEMAC	MH	Unstructured	2D		$S{S}_{wl}$ (9)
[187]	France	Rance estuary	TELEMAC	MH	Unstructured	2D
[188]	France	Seine Estuary	MARS3D	MT	Structured	3D
[189]	Germany	Elbe Estuary	TRIMNP	MH	Structured	2D
[190]	Germany and Germany/Netherlands	Elbe estuary/Jade-Weser estuary/Ems Estuary	UnTRIM	MH	Unstructured	3D
[191]	Germany/Netherlands	Ems Estuary	Non-commercial	MT	-	1D
[192]	Germany/Netherlands	Ems Estuary	Delft3D	MH	Structured	3D
[193]	Germany	Weser estuary	Delft3D	MT	Structured	3D
[194]	Germany/Netherlands	Ems Estuary	SELFE	MH	Unstructured	3D
[195]	India	Narmada Estuary	Non-commercial	MT	Structured	2D
[196]	India	Rushikulya River	Mike	MM	Unstructured	2D
[197]	India	Serayu River	Mike	MM	Unstructured	2D
[198]	India	Ulhas Estuary	Delft3D	MH	Structured	2D
[199]	Indonesia	Jelitik River	Mike	MT	Unstructured	2D
[40]	Indonesia	Kapuas River	Louvain-la-Neuve Ice-ocean Model	MH	Unstructured	2D
[200]	Indonesia	Rokan River Estuary	Mike	MT	Unstructured	2D
[201]	Italy	Misa Estuary	HEC-RAS	MH	-	1D
[202]	Italy	Misa Estuary	Delft3D	MT	Structured	2D
[203]	Malaysia	Pahang River	Mike	MT	-	2D
[204]	Malaysia	Sibu Laut River	DYNHYD5	MH	-	1D
[205]	Morocco	Bouregreg Estuary	HYSED	MM	Structured	2D	MORFAC = 1
[12]	Morocco	Oum-Errabia River	Mike	MT	Unstructured	2D
[206]	Mozambique	Beira Estuary	Delft3D	MH	Structured	2D
[207]	Myanmar	Sittaung Estuary	iRIC	MM	Structured	2D	dt mor = 0.2 s/dt hydro = 1 s
[208]	Netherlands	Rotterdam Waterway	Delft3D	MM	Unstructured	2D	MORFAC = 26
[209]	Netherlands	Scheldt Estuary	Delft3D/Telemac	MH	Structured/Unstructured	2D/3D
[210]	Netherlands	Scheldt Estuary	Delft3D	MM	Structured	2D	MORFAC = 20
[211]	Netherlands	Scheldt Estuary	Non-commercial	MT	Structured	3D
[212]	Netherlands	Scheldt Estuary	TELEMAC	MH	Unstructured	2D
[213]	Netherlands	Western Scheldt estuary	TELEMAC	MT	Unstructured	2D
[214]	New Zealand	Kaipara River	Delft3D	MH	Structured	2D/3D
[215]	New Zealand	Tairua Estuary	Mike	MH	Structured	2D
[216]	Peru	Virrilá Estuary	Delft3D	MT	Structured	2D
[217]	Poland	Kacza river estuary	HEC-RAS	MH	Structured	2D
[218]	Portugal	Douro Estuary	Xbeach	MM	Structured	2D	MORFAC = 10
[219]	Portugal	Douro Estuary/Minho Estuary	Delft3D/Telemac	MH	-	--
[220]	Portugal	Douro Estuary	Delft3D/Telemac	MH	Structured/Unstructured	2D
[221]	Portugal	Douro Estuary	MOHID	MH	-	2D
[222]	Portugal	Lima Estuary	SIMSYS	MH	Structured	2D
[223]	Portugal	Douro Estuary/Minho Estuary	Delft3D/Telemac	MH	Structured/Unstructured	--
[224]	Portugal	Mondego Estuary	Delft3D	MM	Structured	2D
[225]	Portugal	Tagus Estuary	SCHISM	MM	-	2D	MORFAC = 1
[226]	Portugal	Tagus Estuary	SIMSYS	MH	Structured	2D
[227]	Portugal	Tagus Estuary	MOHID	MT	-	2D
[228]	Rusia	Onega Estuary	Delft3D/Hec-Ras	MH	Structured/Structured	1D/2D/3D
[229]	South Africa	Breede Estuary	Delft3D	MH	Structured	2D
[230]	South Korea	Nakdong Estuary	COAWST	MT	Structured	3D
[231]	South Korea	Nakdong Estuary	COAWST	MT	Structured	3D
[232]	South Korea	Nakdong Estuary	CCHE	MM	Structured	2D
[233]	Spain	Guadalquivir Estuary	Delft3D	MT	Structured	2D
[234]	Spain	Oka estuary	SMC	MT	-	2D
[235]	Spain	Ría de Ribadeo	Delft3D	MM	Structured	2D	MORFAC = 12
[236]	Spain	Ría de Viveiro	Delft3D	MH	Structured	2D
[237]	Spain	Ría de Ferrol	ROMS	MH	Structured	3D
[238]	Spain	Suances Estuary	Delft3D	MH	Structured	3D
[239]	Spain	Suances Estuary	H2D	MH	Structured	2D
[41]	Spain/Portugal	Guadiana Estuary	MOHID	MH	-	2D
[240]	Spain/Portugal	Minho Estuary	TELEMAC	MH	Unstructured	2D
[241]	Spain/Portugal	Minho Estuary	Delft3D	MM	Structured	2D	MORFAC = 52/1/1/1
[242]	Taiwan	Beinan Estuary	Mike	MM	Unstructured	2D	dt mor = 3600 s/dt hydro = 3600 s
[243]	Taiwan	Danshui Estuary	SELFE	MH	Unstructured	3D
[244]	Taiwan	Danshui Estuary	Mike	MH	-	1D
[245]	Taiwan	Danshui Estuary	CCHE	MM	Structured	2D	dt mor = 100 s/dt hydro = 10 s
[246]	Taiwan	Lanyan Estuary	Non-commercial	MM	Structured	2D
[247]	Taiwan	Tamsui Estuary	CCHE	MM	Structured	2D	dt mor = 100 s/dt hydro = 10 s
[248]	Taiwan	Tamsui Estuary	CCHE	MM	Non-orthogonal	2D	dt mor = 100 s/dt hydro = 40 s
[249]	Taiwan	Zengwen Estuary	NearCoM-TVD	MM	Structured	3D	MORFAC = 12
[250]	United Kingdom	Taf Estuary	Delft3D	MH	Structured	2D		$S{S}_{vel}$ (1)
[251]	United Kingdom	Avon Estuary	Non-commercial	MT	-	1D
[252]	United Kingdom	Blyth Estuary	TELEMAC	MT	Unstructured	2D
[253]	United Kingdom	Conwy Estuary	CAESAR-Lisflood	MH	-	2D
[254]	United Kingdom	Deben Estuary	Delft3D	MM	Structured	2D	MORFAC = 12
[255]	United Kingdom	Deben Estuary	Delft3D	MM	Structured	2D	MORFAC = 12
[256]	United Kingdom	Dyfi Estuary	TELEMAC	MM	Unstructured	2D	dt mor = 10 s/dt hydro = 10 s
[257]	United Kingdom	Ribble Estuary	Delft3D	MT	Structured	2D
[258]	United Kingdom	Ribble Estuary	Delft3D	MT	Structured	2D
[259]	United Kingdom	Mersey River/Ribble Estuary/Dee River	TELEMAC	MM	Unstructured	2D	dt mor = 600 s/dt hydro = 12 s
[50]	United Kingdom	Severn Estuary	Galerkin model DG-SWEM	MH	Unstructured	2D
[51]	United Kingdom	Severn Estuary	[260]	MH	Unstructured	2D
[14]	United Kingdom	Severn Estuary	Delft3D	MH	Structured	2D
[52]	United Kingdom	Severn Estuary	[260]	MH	Unstructured	2D
[39]	Uruguay/Argentina	Río de la Plata	WQMAP	MH	Structured	2D
[261]	United States	Breton Sound Estuary	FVCOM	MH	Unstructured	3D
[262]	United States	Cape Fear River Estuary	ROMS/WRF-Hydro	MH	Structured	2D
[263]	United States	Chesapeake Bay	COAWST	MT	Structured	3D
[264]	United States	Chesapeake Bay	ROMS	MT	Structured	3D
[16]	United States	Columbia Estuary	Delft3D	MH	Structured	3D
[54]	United States	Columbia Estuary	Delft3D	MH	Structured	2D/3D
[265]	United States	Columbia Estuary	AdH	MH	Unstructured	2D
[21]	United States	Delaware Estuary	Delft3D/Hec-Ras	MH	Structured/Unstructured	1D/2D
[266]	United States	Delaware Estuary	Delft3D/Hec-Ras	MH	Unstructured	1D/2D
[57]	United States	Hudson Estuary	ROMS	MT	-	--
[56]	United States	Hudson Estuary	ROMS	MT	Structured	3D
[55]	United States	Hudson Estuary	ROMS	MT	Structured	3D
[267]	United States	Mobile Bay	ADCIRC	MH	Unstructured	2D
[268]	United States	Saint Johns Estuary	Delft3D	MH	Structured	2D
[269]	United States	San Francisco Bay	Delft3D	MH	Unstructured	2D
[270]	United States	San Francisco Bay	SUNTANS	MT	Unstructured	3D
[271]	United States	Skagit Estuary	FVCOM	MT	Unstructured	3D		$S{S}_{wl} \left(5\right)$ $S{S}_{vel}$ (6)
[272]	United States	Snohomish Estuary	FVCOM	MH	Unstructured	3D
[273]	United States	St. Johns and Nassau Estuary	ADCIRC	MH	Unstructured	2D
[38]	United States	Tillamook Bay	ADCIRC	MH	Unstructured	2D
[274]	United States	Weeks Bay	EFDC	MH	Structured	3D
[275]	Vietnam	Cua Dai Estuary	Delft3D/Mike	MT	Structured	1D/3D
[276]	Vietnam	Mekong Delta	Delft3D	MH	Unstructured	1D/2D		$S{S}_{wl}$ (6)
[277]	Vietnam	Mekong Delta	Delft3D	MT	Structured/Unstructured	2D/3D
[278]	Vietnam	Mekong Delta	Delft3D	MT	Structured	3D		$S{S}_{wl} \left(2\right)$ $S{S}_{vel}$ (2)
[279]	Vietnam	Soai Rap Estuary	TELEMAC	MT	Structured/Unstructured	2D		$S{S}_{wl}$ (4)
[280]	Vietnam	Song Hau channel	Delft3D	MH	Structured	2D/3D
[281]	Vietnam	Thu-Bon Estuary	TELEMAC	MM	Unstructured	2D

.

DELFT3D is a Dutch-origin software developed by Deltares, which is open-source with a paid version alternative. The hydrodynamic module, for both the Flow and Flexible Mesh (FM) versions, solves the Navier–Stokes equation, considering an incompressible fluid under the assumptions of shallow waters and Boussinesq [71,72]. A significant difference between the Flow/FM models lies in the solution method: the Flow version uses finite differences on a structured grid [71], while FM uses finite volumes on an unstructured grid [72]. Both modules have 3D calculation capabilities with a sigma layer or Z-layer in the vertical direction [71,72].

TELEMAC is a free software of French origin, developed by the Laboratoire National d’Hydraulique et Environnement. Its 3D version solves the Reynolds-Averaged Navier–Stokes equations (RANS) for an unstructured horizontal grid [73] and a sigma layer in the vertical direction [74]. The 2D version performs calculations on an unstructured grid, solving the shallow water equations using conservation of momentum and continuity [73,75].

FVCOM is open-source software developed by Chen, H. Liu, and R. C. Beardsley at the University of Georgia [76]. The model is a 3D implementation of the Navier–Stokes equations, where an unstructured grid is solved using finite volumes, with or without the hydrostatic fluid assumption [77,78,79]. For the vertical direction, the tool solves in a sigma layer [77].

Lastly, MIKE is commercial software of Danish origin developed by DHI. In its various versions solve the Navier–Stokes equations under both hydrostatic (MIKE 21 FM) and non-hydrostatic (MIKE 3) conditions. In both cases, the horizontal domain is discretized using unstructured grids, while the vertical direction is discretized through sigma layers [80,81,82].

3.2. Configuration of Hydrodynamic Models

The software packages offer great versatility when configuring the analysis models. Key aspects to consider include the model’s dimension and mesh topology. From a dimensional perspective, three model variants can be distinguished: 1D, 2D, and 3D. Notably, 1D models allow for determining hydraulic characteristics solely in the flow direction. In contrast, 2D models consider both the flow direction and either transverse or vertical dimensions. On the other hand, 3D models provide the capability to represent flow in all directions [42]. It is essential to note that increasing dimensionality involves a higher demand for data inputs and computational resources [83,84].

The literature analysis indicated that authors tend to use depth-averaged 2D models (59.5%), mainly in MH, followed by MM and MT. In second place are 3D models (32.3%), which are most used in MT, followed by MH, and finally MM. Lastly, 1D models are in the third position (8.1%), being used in MH and MT models. It is worth mentioning that for statistical purposes, models combining more than one dimension were considered as different models.

A model’s numerical mesh represents a spatial discretization of the natural environment, which will influence the quality of the results obtained [85]. Currently, there are two types of meshes: structured and unstructured. Structured meshes consist of uniform elements that follow a consistent orientation [86]. Although this type of element yields good results, it does not allow for efficient local refinement [86,87]. On the other hand, unstructured meshes enable a better characterization of complex bathymetries due to their greater flexibility, allowing for more detailed local refinement [86,87,88]. However, they have the disadvantage of losing orthogonality and stretching cells in the flow direction [87].

The literature analysis indicates no clear dominance of one mesh topology over another. The use of structured meshes accounts for 52.8%, mainly employed in MH and MT models. Conversely, unstructured meshes represent 47.2% of the articles, being widely used in MH models and to a lesser extent in MT. Regarding MM models, both mesh types are used, with structured meshes at 10% and unstructured meshes at 7.8% (Figure 7).

By summarizing the information in

Table 2. Results summary.

	MH			MT			MM
	2D	3D	Total	2D	3D	Total	2D	3D	Total
Structured	34%	16%	50%	19%	40%	60%	53%	3%	56%
Unstructured	38%	12%	50%	21%	19%	40%	38%	6%	44%
Total	72%	28%	100%	40%	60%	100%	91%	9%	100%

, it is possible to identify a dependence on topology based on the model type. The analysis reveals distinct preferences in model dimensionality and grid structure across modeling themes. Hydrodynamic (MH) and morphological (MM) applications most commonly employ two-dimensional (2D) models, particularly in configurations aimed at representing surface flow patterns and morphological evolution [89]. In contrast, sediment transport (MT) studies tend to favor three-dimensional (3D) models, especially when simulating suspended sediment dynamics that require vertical resolution [90,91]. These MT applications are frequently implemented using structured meshes, likely due to their stability and compatibility with established transport algorithms in widely used modeling platforms [92].

3.3. Precision of Hydrodynamic Models

The selected studies use different statistical measures to evaluate the precision of the models. One of the most commonly used criteria is the Skill Score (SS), which indicates the correspondence between the model’s results and data obtained from field measurements [93]. Equation (1) represents the SS, where: $X_{model}$ is the simulation result, $X_{obs}$ is the measured information, $\overline{X_{obs}}$ is the average of the measured data. Moriasi et al. (2007) [94] defined different ranges to evaluate model performance using SS: Very Good (1 < SS > 0.75), Good (0.65 < SS < 0.75), Satisfactory (0.5 < SS < 0.65), and Unsatisfactory (SS < 0.5). Although references will depend on the selected author, these ranges allow for comparisons between different models [93].

The SS values of the three most-used software packages, evaluated for both water level and velocity, were extracted directly from the original publications and are presented in Figure 8. Overall, the median of all numerical tools falls within the “Very Good” range. DELFT3D stands out for its precision in calculating water levels, while FVCOM performs best in velocity calculations, both with minimal data dispersion. Although FVCOM also shows a very good performance for water level modeling, this result may be influenced by its smaller sample size (N: 6), which is significantly lower than that of the other models. In contrast, the velocity performance of TELEMAC and DELFT3D is lower compared to the other models, but most measurements still fall within the “Satisfactory” range. While these models may not fully capture the complexity of hydrodynamic phenomena in estuaries, their results generally remain within acceptable limits.

$$SS=1-{\displaystyle \frac{\sum {\left(X_{model}-X_{obs}\right)}^2}{\sum {\left(X_{obs}-\overline{X_{obs}}\right)}^2}}$$

(1)

3.4. Morphological Implementation of Models

The three most used software packages implement a morphological module responsible for updating the bathymetry based on the hydrodynamics and respective sediment transport calculations. DELFT3D updates the bathymetry based on bedload sediment transport mass balance, and a key aspect of it corresponds to the MORFAC parameter, which indicates how many hydrodynamics timesteps will pass before updating the bathymetry [95]. Equation (2) mathematically represents the calculation for the update, where: $∆t$ is the computational timestep, $f_{MORFAC}$ is the morphological acceleration factor, $A^{m,n}$ is the cell area at location $(m,n)$, $S_{b,uu}^{m,n}$ and $S_{b,vv}^{m,n}$ are the bedload sediment transport in the $u$ and $v$ directions, respectively, $∆x^{m,n}$ is the cell width in the $x$ direction, and $∆y^{m,n}$ is the cell width in the $y$ direction.

$${∆}_{SED}^{m,n}={\displaystyle \frac{∆t f_{MORFAC}}{A^{m,n}}}\left(\begin{array}{c}{S}_{b,uu}^{m-1,n}∆y^{m-1,n}-S_{b,uu}^{m,n}∆y^{m,n}+\\ S_{b,vv}^{m,n-1}∆x^{m,n-1}-S_{b,vv}^{m,n}∆x^{m,n}\end{array}\right)$$

(2)

FVCOM implements a community sediment transport model [96], in which bedload fluxes, erosion, and deposition rates are multiplied by a scalar to accelerate the update process [97]. TELEMAC applies a morphological factor that increases the timestep used for morphological simulations [98].

All three models implement a morphological time-scale factor to determine bathymetric changes. As mentioned earlier, two approaches for applying this factor were identified. The first involves the use of a MORFAC value, which ranged from 1 to 26 across studies, with values near or above 50 considered outliers. The second approach defines an update period, which varies between 0.2 s and 100 s, with 3600 s as an outlier (Figure 9).

4. Discussion

The thematic scope of the review was intentionally shaped by a filtering strategy designed to identify studies that apply numerical models to simulate estuarine morphodynamics. The findings align closely with the study’s stated objectives and offer a coherent synthesis of current practices in estuarine modeling. One of the primary contributions is the identification of Delft3D, TELEMAC, and FVCOM as the most commonly used software tools, which directly addresses the objective of evaluating model prevalence in estuarine environments. However, this result must be interpreted within the context of the review’s thematic scope and inclusion criteria. The keyword-based filtering strategy prioritized studies focused on morphodynamic evolution and this emphasis does not indicate methodological bias but rather mirrors the structure of the literature and the specific aims of the review.

Recognizing the interdependence of morphological evolution with hydrodynamic and sediment transport processes, the search also included studies classified under hydrodynamics (MH) and sediment transport (MT), provided they were clearly framed within a morphodynamic modeling (MM) context. Hydrodynamic modeling allows for the identification of flow changes due to environmental factors [99,100,101,102], while sediment transport and morphological models help identify the transformations experienced by the study system as a result of hydrodynamic variations [103,104,105,106]. While this approach ensured alignment with the study’s objectives, it may have inadvertently excluded relevant works that addressed similar processes using different terminology in titles or abstracts—a common trade-off in systematic reviews between thematic focus and broader inclusiveness [107].

The thematic distribution observed—where hydrodynamic models appear more frequently than sediment transport and morphological models—should also be interpreted in light of both the applied filtering criteria and the inherent structure of numerical modeling in estuarine environments. Hydrodynamic modeling serves as the foundational layer upon which sediment transport and morphological simulations are constructed, as variables such as water depth and velocity fields are prerequisites for calculating sediment fluxes and bed evolution. As such, every morphodynamic model inherently includes a hydrodynamic component, and many studies prioritize hydrodynamic simulation before integrating additional processes. This prevalence is further shaped by the conservative and hierarchical classification framework used, where studies were categorized according to the highest level of process integration. While this ensured thematic clarity, it may have underrepresented the extent of process coupling in multi-component models. Therefore, the predominance of hydrodynamic models in the results reflects methodological decisions and the foundational role of hydrodynamics rather than a conceptual imbalance in the field.

Beyond model prevalence, this review also examined configuration strategies across studies. The analysis of dimensionality, mesh topology, and the use of morphological acceleration factors provided insight into how model setups are adapted to estuarine conditions, revealing prevailing practices and the influence of both system characteristics and computational considerations in model design. Although the review systematically categorized modeling tools by mesh type, model dimensionality, and morphological coupling, it did not extend this classification to external forcing mechanisms (e.g., wave action, wind stress) or internal numerical configurations (e.g., turbulence closure schemes, wetting–drying algorithms). These elements are highly scenario-specific and tied to the physical and operational context of each study. Moreover, they are not consistently reported in the literature, and any attempt at generalization would risk misinterpretation regarding model completeness or performance. In particular, while the inclusion of the morphological acceleration factor (MORFAC) offers insight into current practices, its application is rarely justified in the reviewed literature. For this reason, its presence in the analysis should be interpreted with caution and understood as reflecting reporting tendencies rather than an endorsement of its appropriateness in all modeling contexts.

Regarding model validation, all Skill Score (SS) values were extracted directly from the original publications without recalculation or adjustment. To account for differences in validation targets—particularly between water level and velocity—the SS analysis was disaggregated by variable to ensure meaningful comparisons. This approach enabled the identification of performance trends across modeling tools while minimizing the impact of heterogeneous validation metrics. Notably, all reported SS values fell within the range commonly interpreted as “good” or “very good”, suggesting a general consistency in performance regardless of platform. As such, SS outcomes did not serve as a decisive factor in comparing model capabilities but reflected the widespread adoption of similar validation practices.

In contrast to the relatively well-documented validation of hydrodynamic modules, performance assessment of sediment transport and morphological evolution was less consistently reported. While many studies simulated both suspended and bedload transport, the observed emphasis on bedload reflects how processes were described in the original publications rather than a review bias. Bedload is often highlighted as a dominant mechanism in short-term morphological studies, whereas suspended transport, though commonly modeled, is less frequently validated or discussed in detail. Morphological validation, when present, tended to be qualitative—based on erosion–deposition patterns or bathymetric comparisons—rather than supported by standardized quantitative metrics. This underscores the need for more consistent documentation and standardization of validation approaches in sediment and morphodynamic modeling.

Finally, the review identified thematic and geographic trends that further support its broader aim of characterizing the current state of estuarine modeling. For instance, there was a notable emphasis on hydrodynamics over morphological modeling and a dominance of case studies conducted in Asia. Additionally, the review highlighted trade-offs that influence model selection in practice. Structured-mesh tools like Delft3D are widely used due to their integration of hydrodynamic and morphological modules, user-friendly graphical interfaces, and open-source accessibility. However, they may present limitations in spatial flexibility and demand higher computational resources, particularly in 3D applications [22]. In contrast, unstructured-mesh models such as TELEMAC and FVCOM offer superior spatial adaptability and better representation of complex bathymetries but often require greater technical expertise and processing power [108,109]. Proprietary platforms provide strong technical support and ease of use, though access may be restricted in resource-constrained locations. These factors help explain the observed patterns in model usage, illustrating that model selection is shaped not only by technical performance but also by project-specific objectives, user experience, and institutional access.

5. Conclusions

This systematic review analyzed 197 peer-reviewed studies published over the past 15 years that applied numerical models to estuarine systems for the simulation of morphodynamic evolution, sediment transport, and hydrodynamic behavior. These three modeling practices were examined jointly to support a comprehensive understanding of morphodynamic modeling efforts, which inherently depend on the accurate representation of both sediment transport processes and hydrodynamic forcing. The analysis identified the most frequently used software platforms, described common configuration practices, and examined validation approaches. Results show a predominance of 2D depth-averaged models and structured meshes, with Delft3D, TELEMAC, and FVCOM being the most widely used tools. Validation was most commonly reported for water level and velocity, with Skill Scores typically falling in the ‘good’ to ‘very good’ range; however, sediment transport and morphological validation remain inconsistently reported. The review also identified geographic imbalances in model applications and emphasized the importance of scenario-specific forcing and solver configurations. Future work should prioritize improved reporting of sediment dynamics, performance metrics for morphology, and comparative analyses of model capability across estuarine typologies. These efforts would enhance reproducibility, support better-informed model selection, and foster greater standardization in estuarine hydro-morphodynamic modeling practices.