Applications of Vulnerability Assessment and Numerical Modelling for Seawater Intrusion in Coastal Aquifers: An Overview

Abstract

Seawater intrusion forms a significant environmental and hydrogeological phenomenon that raises significant risks for the sustainability and quality of coastal aquifer hydrosystems. The present review study critically examines the available methodologies for assessing aquifer susceptibility to seawater intrusion, including the GALDIT and SEAWAT models. The GALDIT model is a parametric model that uses six main hydrogeological parameters for assessing groundwater vulnerability to seawater intrusion. Numerous researchers have proposed improvements to GALDIT either by adding new variables such as well density, well pumping rates, and hydrochemical indicators, or by applying machine learning (ML), fuzzy logic, and optimization algorithms to improve spatial resolution and accuracy. The SEAWAT code can be used for simulating variable-density groundwater flow and solute transport and has been widely used to model the salinization process under different pumping and sea-level rise scenarios. The presented case studies show that the combination of GALDIT and SEAWAT offers a stronger and robust framework for both vulnerability zoning and dynamic flow and transport simulation. Recent SEAWAT studies show that paleo-salinization has a significant influence, highlighting the need to measure both the trapped saline water in confined layers and the lateral intrusion of seawater. The present review concludes that future efforts need to focus on hybrid modeling approaches, integration of hydrochemical and geophysical data, and the inclusion of anthropogenic and climate-associated factors to enhance the accuracy and applicability of seawater intrusion risk assessments in coastal areas.

1. Introduction

The most common issue in Mediterranean coastal aquifers is seawater intrusion, which contaminates freshwater and decreases its availability for uses like irrigation [1,2,3,4]. Intrinsic vulnerability is typically regarded as a characteristic of groundwater, reflecting its susceptibility to both human and natural influences [5]. In many Mediterranean areas, reduced recharge during extended dry periods increases the risk of seawater intrusion, highlighting the need for vulnerability tools such as GALDIT, and using models like SEAWAT for additional assessment [6,7]. Furthermore, increased water needs have led to deeper drilling as water quantities in shallow layers have decreased, which increases pumping rates and accelerates the propagation of the seawater intrusion front, making coastal aquifers even more vulnerable to salinization. Even a small amount of saltwater mixing with freshwater can render it unsuitable for any use, potentially leading to the need for alternative sources [8]. The mixing of seawater into freshwater aquifers is a complex process influenced by hydrogeochemical reactions, shoreline geomorphology, biological processes, aquifer flow, and other factors [9]. Specific ions such as Cl, Na, Mg, SO₄, and Br become enriched due to seawater intrusion and can be used as indicators [10,11,12]. Vulnerability mapping helps with successful groundwater management by providing important data on aquifer conditions [4,13]. Understanding the groundwater flow in coastal aquifers is essential, as local pumping and hydraulic gradient directly control the movement of the saltwater, determining whether vulnerability monitoring tools are sufficient or whether detailed numerical simulations are required [1,14]. Geochemical tracers such as Cl⁻ and water isotopes (d²H, d¹⁸O, ³H) have also been widely used to evaluate seawater intrusion from mass-balance mixing calculations [15,16]. Additionally, index-based methods, such as the Water Quality Index and Seawater Mixing Index, are commonly used to identify seawater intrusion in coastal aquifers [16,17]. The Water Quality Index combines essential hydrogeochemical parameters such as pH, electrical conductivity (EC), total dissolved solids (TDS), chloride (Cl⁻), sulfate (SO₄²⁻), nitrate (NO₃⁻), phosphate (PO₄³⁻), sodium (Na⁺), and calcium (Ca²⁺) into one dimensionless qualtnity for overall groundwater quality [18]. Field and GIS (Geographic Information System)-based approaches also became widely adopted, especially with the development of methods such as the GALDIT vulnerability mapping technique [19]. Another method to delineate areas more vulnerable to seawater intrusion is the AVI (Aquifer Vulnerability Index) method, which uses two parameters: thickness of each overlying sedimentary layer above the uppermost saturated aquifer surface (d) and estimated hydraulic conductivity (k) [20]. Similarly, Azizi et al. [6] proposed the ILDRT method, which evaluates groundwater vulnerability using five hydrogeological parameters: the impact of existing saltwater intrusion, groundwater level above mean sea level, distance from the salinity source, recharge, and aquifer thickness.

Several indexes have been proposed for assessing aquifer vulnerability, including DRASTIC [21], GALDIT [19], EPIK [22], PI [23], and SVAP [24]. The most commonly used and effective parametric models for determining groundwater vulnerability are DRASTIC for nitrate contamination and GALDIT for seawater intrusion, despite certain drawbacks [25,26]. In cases where datasets are limited, these challenges can be addressed by employing advanced data fusion techniques and comparative analyses using machine learning and deep learning approaches [27]. Such models have been used for a variety of purposes, including simulating groundwater flow, predicting water quality, modeling seawater mixing, assessing vulnerability to nitrate pollution, and estimating groundwater levels [28,29].

For GALDIT, according to Chachadi and Lobo Ferreira, [19], there are six factors, including groundwater occurrence (G), aquifer hydraulic conductivity (A), height of groundwater level above sea level (L), distance from the shoreline (D), impact of existing saltwater intrusion (I), and thickness of the aquifer (T). Based on the results, the height of groundwater above sea level and distance from the shore, both weighted 4, are critical, as they control the hydraulic gradient affecting seawater intrusion. The GALDIT method has been used by numerous authors, some of them have modified it by adding new parameters or by changing some of the ones that already exist.

For hydrogeologists, numerical modeling is an effective tool that aids in understanding the process of seawater intrusion and estimating seawater contamination in coastal aquifers [30]. SEAWAT is one of the most widely used numerical codes for simulating variable-density groundwater flow and assessing seawater intrusion in coastal aquifers [31,32]. However, although numerical simulations can be reliable and useful, they require a lot of hydrogeological data as inputs, have a demanding calibration process, and require a lot of processing power for data curation and modeling [33]. In areas that have faults and fractured formations, SEAWAT simulations are often supported and greatly assisted by geophysical surveys to better characterize groundwater salinity sources [34]. These surveys help constrain key boundary conditions and hydraulic properties, improving the reliability of the numerical results.

The objective of this study is to provide a comprehensive summary of the GALDIT vulnerability index and the SEAWAT numerical model applications in scientific literature. The overview aims to highlight each approach addresses seawater intrusion in coastal aquifers and describe their effectiveness in supporting sustainable groundwater management. This review does not treat GALDIT and SEAWAT as similar methods or interchangeable alternatives. Instead, it assesses how a parametric screening tool (GALDIT) and a physics-based simulation model (SEAWAT) can be effectively combined.

2. Methods for Assessing Seawater Intrusion Processes in Coastal Aquifers

To provide an overview of the methodological framework outlined in this review, Figure 1a,b illustrates the general workflow of the GALDIT method for coastal aquifer vulnerability mapping (Figure 1a) and the main steps of the SEAWAT numerical code for seawater intrusion simulation (Figure 1b). The schematic shows the specific steps necessary to implement each method, GALDIT for vulnerability mapping and SEAWAT for numerical simulation of seawater intrusion.

3. Fieldwork and Analytical Techniques

3.1. Hydrogeological Conditions and Investigations

Unconfined aquifers are typically more susceptible to seawater intrusion compared to confined systems. Also, sedimentary aquifers, which constitute the primary groundwater reservoirs, are experiencing growing impacts from seawater intrusion [6,8,35]. Groundwater level measurements, drill logs, and groundwater survey reports are able to contribute the most critical data, while analysis of pumping test data provide essential insights into aquifer characteristics, including transmissivity, storage capacity, and hydraulic conductivity [7]. The above data are essential and needed for the methods under investigation and, specifically in GALDIT, they are used as parameters for mapping vulnerable zones through GIS overlays, while in SEAWAT, they control the model calibration and numerical simulation of density-dependent flow. Therefore, a significant amount of hydrogeological data help in predictive modeling of seawater intrusion [36]. In addition, in coastal karstic aquifers, vulnerability assessment cannot be based solely on groundwater level, pumping test data and hydrochemistry. As Kresic & Panday (2020) [36] suggest, flow is simultaneously governed by the rock matrix, faults and conduits, requiring data on structural discontinuities, degree of karstification, conduit geometry and spring discharge. Recharge in karstic systems is constantly changing and depends largely on rainfall, infiltration through sinkholes, and epikarst storage, making accurate rainfall records fundamental to any assessment of groundwater vulnerability in this type of aquifers.

3.2. Geoelectrical Methods for Seawater Intrusion

Geoelectrical techniques have proved to be especially useful for mapping and identifying seawater intrusion, in combination with the field methods used in GALDIT and SEAWAT applications. Lower resistivity usually means higher salinity, which helps scientists understand the extend of seawater intrusion withing an area [4,6]. Multi-electrode resistivity surveys are frequently used in several areas to detect and spatially map salinity fronts in coastal aquifers. In particular, the resistivity measurements provide valuable field data for validating the vulnerability components used in GALDIT. In addition, although seismic methods do not directly detect salinity, they can yield structural information that is necessary to define the conditions controlling groundwater flow in each area, providing the necessary data for SEAWAT models. Therefore, these techniques are tools that contribute to the accuracy of the subsurface structure and the spatial distribution of saltwater. In addition to surface geophysical methods, borehole geophysical logs, such as natural gamma ray and electrical resistivity loggings, are important tools that can help to identify geological formations, offering more reliable differentiation between fresh and brackish water compared to surface methods. For example, in the study conducted in Lagos, Nigeria, the integration of natural gamma ray and resistivity logs helped to map the fresh-salt boundaries and develop the conceptual model of the subsurface system [37].

3.3. Chemical Analysis

3.3.1. Chemical Parameters

Numerous investigations have been carried out to measure the extent of seawater intrusion in study areas by using the geochemical and isotopic characteristics of groundwater from unconsolidated sedimentary aquifer systems [18,38]. Salinity, chloride concentration, and other chemical indicators of seawater incursion are typically measured in groundwater samples. Elements such as NO₃⁻, F⁻, HCO₃⁻, and PO₄³⁻ are assigned lower weights, as they have minor role in the mixing processes between freshwater and seawater. In contrast, elements like Na⁺, Ca²⁺, Cl⁻, and SO₄²⁻ are given greater weight due to their significant involvement in the mechanisms of seawater intrusion [1,18]. Within the GALDIT method, it is recommended using the Cl⁻/(HCO₃⁻ + CO₃²⁻) ratio to assess the extent of seawater intrusion in coastal aquifers, with a threshold concentration ranging from 40 to 300 mg/L Cl⁻ considered as indicative of saltwater intrusion [6,39].

3.3.2. Isotope Analyses

Isotopic tracers are used in several studies because they provide information on groundwater quality that cannot be obtained solely through conventional hydrochemical analyses. Deuterium and oxygen-18 are two examples of isotopes that are used to determine the origin of water sources and identify the types of mixing between freshwater and saltwater. In addition, strontium isotopes are particularly useful for identifying water-rock interaction processes, as the ⁸⁷Sr/⁸⁶Sr ratios remain stable during most geochemical reactions and are largely controlled by the lithology of the aquifer material. This allows Sr isotopes to act as reliable tracers for distinguishing modern seawater intrusion from the contribution of deeper palaeo-saline waters and salinity controlled by the lithology of the aquifer, consequently explaining the source of groundwater salinization. Therefore, different minerals can provide isotopic signatures, allowing the distinction between different sources and mixing processes within aquifers.

3.4. Monitoring Systems

Sustainable groundwater resources management requires deep understanding of how hydraulic heads fluctuate [2]. Periodic changes in salinity, groundwater quality, and groundwater levels are typically monitored through a network of observation wells. These installations may include data loggers and multiparameter probes capable of recording temperature, electrical conductivity, TDS, and water table depth at high-frequency intervals, although real-time measurements require telemetry systems, which are not always available [40]. Additional monitoring data should include spring discharges of karst systems as well as rainfall data, as they highly affect groundwater recharge and are therefore essential for any hydrogeological assessment or modeling application. An effective groundwater-monitoring network should cover the aquifer sufficiently, include both coastal and inland monitoring points, and where possible measure groundwater conditions at different depths to detect vertical changes in salinity. Also, it should combine continuous automated measurements through frequent manually sampling to collect both extensive time-series data and accurate chemical information. However, factors such as high installation and maintenance costs, limited equipment, difficult access to certain sites, and the need for telemetry systems often make it challenging to develop a fully optimized monitoring network.

4. The GALDIT Method

4.1. Description and Application

The GALDIT method assesses aquifer vulnerability to seawater intrusion by considering the hydrogeological conditions that primarily influence the potential for seawater intrusion in the area [4,6]. The calculation of the GALDIT index, along with the combination of many parameter layers in a Geographic Information System (GIS), allows the identification of critical areas that are significant and vulnerable to seawater incursion [40]. The GALDIT data layers encompass the hydrogeological and hydrochemical characteristics of the aquifer and include: Groundwater occurrence (G), Aquifer hydraulic conductivity (A), Groundwater level height (L), Distance from the shore (D), Impact of existing seawater intrusion status (I), and Aquifer thickness (T).

From these, the GALDIT index is calculated according to Equation (1).

GALDIT index = G_RG_W + A_RA_W + L_RL_W + D_RD_W + I_RI_W + T_RT_W

(1)

In the GALDIT index, each parameter is assigned a score (R) and a weight (W), where the scores represent the degree of vulnerability based on the field values and the weights express the relative importance of each parameter. Thus, the final vulnerability map that is produced is directly influenced by these two values, with several studies proposing other approaches to the weights depending on local conditions in order to improve the accuracy of vulnerability assessments [41,42]. To illustrate the broad geographical applicability and field practicality of the GALDIT approach, Figure 2 presents a global map with representative study areas where both the traditional GALDIT index and its various modified forms have been utilized to assess the vulnerability of coastal aquifers to seawater intrusion. To present literature trends in the development of modified GALDIT approaches and highlight research collaborations worldwide, a bibliometric network of published studies was created using VOSviewer 1.6.20 (Figure 3). The network depicts groups of authors working on GALDIT variants and hybrid models, showing how the method has developed under alternative environmental and data conditions.

4.2. Parameter-Modified Variants of GALDIT

The GAPDIT method is a different version of GALDIT, which was developed to assess the sensitivity of coastal aquifers to seawater intrusion, replacing the parameter L (height of the aquifer above sea level) with the parameter P (pumping rate) [39]. The GAPDIT index showed higher predictive accuracy than the GALDIT method, particularly in areas with high pumping rates. The method is based on the GALDIT framework and incorporates two adapted variants, GAiDIT and GALDIT-i [43]. These extensions were developed to improve the representation of hydrodynamic conditions in areas where the original GALDIT parameters may not be sufficient. The main modification introduced in GAiDIT concerns the replacement of the parameter L (groundwater level above sea level) with the hydraulic gradient (i). This replacement offers a more realistic representation of transient, pumping-controlled flow conditions, particularly in intensively exploited coastal aquifers. In contrast with the GALDIT-i method, an additional parameter of the hydraulic gradient (i) was added, representing the dynamics of flow towards the sea [42]. The incorporation of the hydraulic gradient towards the sea (i) strengthens the connection with the groundwater quality data (e.g., EC, TDS), and the gradient causes seawater to move directly into the aquifer. Therefore, it has been observed that have stronger seaward hydraulic gradient results in higher salinity, allowing the modified indicators to give better results. GALDIT-i showed improved precision compared to the traditional model, but GAiDIT was considered to be more advantageous in regions with significant human activities.

A modified version of the GALDIT method, known as GALDITE, was used in the study of coastal aquifers in Espírito Santo, Brazil [44]. A new methodology used to determine the amount of water pumped from aquifers assists in incorporating the groundwater pumping when considering groundwater vulnerability to seawater intrusion. The E parameter (groundwater recharge with extraction) was calculated through the water balance, and areas where pumping exceeds natural recharge, resulting in a negative balance, were assigned higher vulnerability values due to their greater susceptibility to seawater intrusion.

Another study, which was carried out in central Iran’s Qom-Kahak region [45] created a modified GALDIT methodology known as GALDITMW, adding two more parameters: M (aquifer porous medium) and W (production wells’ density). The W parameter displays the wells’ concentration in a specific area where seawater intrusion can potentially be more intense, while the M parameter characterizes the grain size.

The research on coastal aquifers in Odisha, India, applied a modified GALDIT method referred to as M-GALDIT, in combination with the Analytic Hierarchy Process (AHP) [46]. According to the modified index, the parameter I (Impact of Existing Seawater Intrusion) was replaced with the GQISWI (Groundwater Quality Index for Seawater Intrusion). The GQISWI is calculated independently as a separate groundwater-quality indicator that quantifies the degree of seawater mixing based on major ion chemistry. Also, it is integrated into the M-GALDIT framework in place of the original Cl/(HCO₃ + CO₃) ratio, effectively adding an additional evaluation layer that improves the representation of seawater intrusion within the vulnerability index.

Overall, the alternatives presented highlight the adaptability of the GALDIT framework. While the inclusion of new parameters improves local accuracy, it also reduces transferability, showing that GALDIT modifications need careful replication to regional conditions. To better illustrate how each modification affects seawater intrusion assessment, their main attributes are presented below:

• GAPDIT replaces L with P (pumping rate), more accurate in regions with intensive groundwater extraction [38].
• GALDIT-i adds hydraulic gradient (i), improving correlation with EC and TDS and enhancing the link to water quality data [42].
• GAiDIT replaces the parameter L (groundwater level above sea level) with the hydraulic gradient i, improving performance in heavily pumped coastal aquifers [42].
• GALDITE incorporates parameter E (groundwater exploitation), incorporating the effect of over-pumping [43].
• GALDITMW adds M (porous medium/grain size) and W (well density), relating vulnerability to aquifer texture and well distribution [44].
• M-GALDIT replaces parameter I with groundwater quality index (GQISWI), making it appropriate when intensive monitoring data are available [45].

4.3. Dynamic and Temporal Approaches to GALDIT

A main attribute of the original GALDIT index is static, since the parameter values do not change over time and we get a picture based on the data that was entered. However, the vulnerability of the aquifer is strongly influenced by the temporal changes in groundwater levels, pumping and seawater intrusion. For this reason, modified approaches have introduced temporal dynamics into the index. A modified GALDIT method classified the parameters as static or dynamic [7]. Static parameters include groundwater occurrence (G), aquifer hydraulic conductivity (A) and distance from the coastline (D), while dynamic parameters include groundwater level (L), the effect of seawater intrusion (I), and aquifer thickness (T). The incorporation of time-dependent variables allows the method to capture seasonal changes and improves the reliability of vulnerability assessments, as the effect of seawater intrusion (I) is quantified through the Revelle Index, a hydrochemical indicator of groundwater salinization in coastal aquifers. Using ten-year monthly averages allowed seasonal changes to be recognized more clearly. The modified method also refined the scoring system from 4 to 10 categories, improving the spatial resolution of vulnerability assessments. Seawater intrusion is also projected to occur by 2050 [47]. Results showed that declining groundwater levels, rising chloride concentrations, and continued over-pumping increased the seawater intrusion risk.

4.4. Optimization and Machine Learning Approaches for GALDIT

One limitation of the original GALDIT method is that it uses fixed parameter weights, which cannot reflect differences between local hydrogeological conditions. Therefore, the contribution of each factor to seawater intrusion may not be represented equally well in all aquifers. To overcome this limitation, recent studies have introduced optimization algorithms and machine learning models to calibrate parameter weights using observed salinity indices and improve the predictive reliability of vulnerability mapping. In Morocco, the GALDIT index was modified to better reflect the specific hydrogeological conditions of the Souss–Massa aquifer [48]. Three new parameters were introduced: river contribution, hydraulic gradient, and probable seawater intrusion, which influence salinity patterns in this area. The final weights of the parameters were not taken from the original GALDIT but were recalculated using AHP and single-parameter sensitivity analysis, allowing the index to be calibrated to local data and to produce more accurate vulnerability results.

4.5. Coupling with Process-Based Models

The integration of GALDIT with physically based numerical models does not imply that the two methods have the same role. Rather, each approach contributes complementary information: GALDIT provides a spatially explicit vulnerability assessment, while numerical models simulate the underlying flow and transport processes. When combined, they offer a more complete framework for interpreting and applying the results. GALDIT quickly delineates areas that are most vulnerable to seawater intrusion under current conditions, acting as a screening tool. GALDIT provides a rapid screening of areas that are more vulnerable to seawater intrusion under current conditions. These identified hotspots can then be examined in greater detail using numerical models such as SEAWAT and MODFLOW, which simulate density-dependent flow and salinity transport to assess future intrusion pathways and evaluate management scenarios.

4.6. Hybrid Approaches Combining GALDIT with Other Indices

The parameter modifications do not change the original GALDIT indicators, but instead incorporate new parameters related to the specific conditions in the study areas. Therefore, when GALDIT is combined with complementary tools such as the Aquifer Vulnerability Index (AVI), which evaluates vertical pollutant transport, or the Water Quality Index (WQI), which summarizes overall groundwater quality, the resulting framework provides a more comprehensive risk assessment that accounts for multiple sources of groundwater salinization, including surface contamination, river intrusion, and anthropogenic impacts. The use of both the GALDIT and ILDRT methods in Lake Urmia, Iran, has shown that the combination of hydrochemical indicators can enhance the final vulnerability map [6]. In addition, the GALDIT-SUSI model incorporates influences from surface waters such as rivers, wetlands, and lagoons, further improving the assessment of groundwater salinization compared to the traditional GALDIT index [49]. These hybrid frameworks highlight a growing trend of extending vulnerability indicators with the intention to take into account additional natural processes that cannot be well represented by the original GALDIT.

4.7. Multi-Criteria and Risk Extensions of GALDIT

Unlike the parameter-based and hybrid variants of GALDIT discussed in previous sections, the multi-criteria and risk-based alternatives are designed not only to represent vulnerability levels but also to support the decision makers by incorporating uncertainty, weighting optimization, and risk prioritization tools. Using approaches such as fuzzy classification, this extension transforms GALDIT into a framework that can assess and identify only actually vulnerable areas. In eastern Tunisia, the GALDIT method was modified by including a Land Use parameter and using sensitivity analysis to optimize and enhance parameter weights, resulting in the creation of the Global Risk Index [25], with results showing that groundwater depth remained the most critical parameter. In northern Greece, the GALDIT-F model was suggested, using fuzzy multi-criteria categorization to make vulnerability mapping more accurate [50]. These extensions show the way GALDIT has developed into a flexible and adaptable framework that can include many different risk factors and criteria. However, it still needs to be carefully modified depending on local conditions. Representative applications of the GALDIT method and its modified versions are summarized in Table S1 (Supplementary Materials).

5. Numerical Modeling with SEAWAT

To complement the detailed description of SEAWAT’s numerical modeling capabilities, this section illustrates how the model is applied in coastal aquifer studies and how its results support vulnerability assessments. To highlight the scientific importance and global use of SEAWAT, a bibliometric co-authorship analysis was performed using VOSviewer 1.6.20 (Figure 4). The network visualizes the major author clusters, active research groups, and international collaborations working specifically on SEAWAT applications in density-dependent groundwater flow modeling. By placing SEAWAT within its broader research context, the analysis demonstrates its widespread adoption across diverse hydrogeological settings and clarifies why SEAWAT is included as a key component of this review. This analysis has been created to place SEAWAT within global research activity, not to compare it with GALDIT.

5.1. SEAWAT Model Description Applications in Case Studies

SEAWAT is a process-based numerical code that simulates variable-density groundwater flow and solute transport by linking MODFLOW with MT3DMS. It solves the flow equations governing density-dependent flow and solute transport, providing results for the extent of the brackish zone as well as mixing processes and interface evolution under different pumping and sea level rise scenarios. Figure 5 illustrates the coastal areas worldwide where SEAWAT has been applied, presenting different hydrogeological environments and water resources management issues.

The map shows how flexible and reliable SEAWAT is as a process-based simulator for challenging conditions such as seawater intrusion, groundwater recharge, pumping, and sea-level rise scenarios. This provides the potential for the protection and management of coastal groundwater. According to the application studies of the SEAWAT model, it has been shown to be suitable for various hydrogeological and climatic environments.

5.2. Main Mechanisms and Findings

SEAWAT has been applied in numerous coastal aquifers to evaluate how salinization evolves under different hydrological and pumping conditions. In the Biscayne aquifer (Florida), groundwater levels, TDS measurements, and pumping data were used for model calibration, and simulations demonstrated that intensified pumping accelerates inland movement of the salinity front [45]. Comparable findings have been reported for Chalkidiki (Greece), where SEAWAT revealed that increased groundwater withdrawals in shallow aquifers enhance chloride and TDS enrichment [51,52]. In larger systems such as the Nile Delta, the model has been instrumental in disentangling the combined effects of groundwater table decline, reduced river flows, and sea-level rise, all of which contribute to seawater intrusion [53,54,55]. Similar applications in the Dago River basin have shown that rainfall patterns and aquifer permeability strongly influence intrusion dynamics [56,57], while studies in the Mekong Delta demonstrated that reduced river levels allow saline water to propagate inland [46]. In the Barreiras aquifer (Brazil), SEAWAT combined with porosity measurements, pumping tests, and geoelectrical tomography showed that faults act as preferential pathways for seawater migration [33]. Collectively, these applications highlight that SEAWAT is most useful as a predictive tool, providing insights into the evolution of coastal aquifers under future stress information that cannot be obtained from static vulnerability indices. Overall, these studies demonstrate that SEAWAT provides a reliable way to estimate where and how intrusion will move in the future and evaluate long-term groundwater management strategies under changing climatic and pumping conditions.

5.3. Management Measures

SEAWAT is an increasingly used tool not only for characterizing seawater intrusion processes, but also for properly evaluating scenario-based simulations. Thus, rather than reporting management outcomes, a model allows users to predict the long-term effectiveness, spatial extent, and economic feasibility of control measures. For example, in Chalkidiki, Greece, SEAWAT simulations compared multiple injection barrier configurations under varying recharge conditions and identified the option with the lowest injection volume as the most cost-effective solution [52]. Similarly, in the Biscayne aquifer in Florida, SEAWAT was used to evaluate mixed physical barriers, showing significant reductions in seawater intrusion [51]. In coastal aquifers, SEAWAT simulations have provided important results on whether relocating wells further away from the coast or reducing groundwater extraction can effectively limit future saltwater intrusion phenomena. For example, in the Amol-Ghaemshahr aquifer in Iran, simulations showed that relocating wells inland could significantly reduce the risk of seawater intrusion [58]. Also, Managed Aquifer Recharge (MAR) applications have been evaluated with SEAWAT, such as in the Ninh Thuan Basin in Vietnam, where artificial recharge scenarios have been shown to be able to stabilize groundwater levels and improve long-term water availability [59]. According to the above examples of application of SEAWAT, researchers can support their future decisions, resulting in the improvement and sustainable management of groundwater.

5.4. Uncertainties and Limitations

The degree of uncertainty in hydrogeological input data, boundary conditions, and pumping rates has a significant impact on how effectively SEAWAT determines events. This is especially true because density-driven flow makes errors incorporated in hydraulic gradients and solute transport more serious. In general, SEAWAT has the ability to simulate density-driven flow and solute transport, but several researchers consider SEAWAT to be sensitive to uncertainties in input data and boundary conditions. It has been reported in several studies that small variations in hydraulic parameters or pumping rates can significantly alter the extent of the salinity front [60,61]. At Ain Sukhna (Egypt), the calibration of the model was limited by the lack of long-term monitoring records, while structural uncertainties associated with faults further affected the reliability of the model [60]. Furthermore, simulations conducted in Laizhou Bay (China) have shown that there is sufficient uncertainty due to geological heterogeneity and combined pumping rates that strongly influence the final model [61]. The need for continuous groundwater and salinity monitoring to support calibration has also been highlighted in Chalkidiki, Greece [52], and in the Nile Delta, where sea-level rise and reduced river discharge scenarios significantly affect seawater intrusion processes [54]. Accordingly, studies have shown that the reliability of SEAWAT depends largely on the input data, the pumping conditions, and the geological characteristics of the area.

6. Advances and Suggested Improvements to the GALDIT Method

6.1. Methodological Extensions

Recent studies that have utilized the GALDIT index highlight its weaknesses in terms of the absence of dynamic parameters (e.g., groundwater pumping and hydraulic gradients), and because it uses the same weights in all areas, it cannot be easily applied to different hydrogeological environments without adaptation. The modified versions, GAiDIT and GALDIT-i, introduce the hydraulic gradient to show how important the flow towards the sea is [43]. In addition, the new modifications of the GALDIT index are more accurate, but they can only be applied where sufficient data is available, and the calibration focuses on the specific research area, which has the consequence that they cannot be automatically applied anywhere. The integration of numerical models (e.g., with MODFLOW) in the GALDIT index has been proposed in order to compensate for the static presentation of the index with the addition of the simulation of temporal changes. However, such an integration of temporal changes in the model requires hydrogeological data, making it impractical in many areas where data are limited. In addition, the multi-criteria fuzzy analysis method can give more adaptable and realistic results, but there is no rule for how it should be used, so each researcher applies it in a different way. Also, each region has different characteristics that may have a different impact on parameters of the GALDIT index, which must be adjusted separately for each region. In addition, the weights of each parameter are given based on the researcher’s judgment and experience, so there is no rule for this. Moreover, the GALDIT index should be recalculated whenever new hydrogeological or monitoring data become available, so that the vulnerability assessment reflects current conditions rather than remaining static. Such developments will allow GALDIT to evolve from a static, locally bound indicator to a more generalized decision-supporting tool for coastal aquifer management.

6.2. Climate and Dynamic Factors

Climate change, which is increasingly affecting more areas with extended periods of droughts, reduced recharge, and sea level rise, continuously changes the dynamics of seawater intrusion into coastal aquifers. According to the GALDIT index, the parameters are treated as static, considering that groundwater levels, recharge conditions, and pumping rates remain unchanged over time. This assumption is not valid for longer time periods and can lead to incorrect classification of vulnerability, especially in areas where salinization is occurring rapidly. In general, studies that have applied GALDIT report that the weaknesses of the index are not related to its structure, but to its inability to update its scores and weights when hydrological pressures evolve. Also, the ratings and weights of parameters should be periodically recalculated based on seasonal or long-term monitoring data to reflect the evolution of seawater intrusion in real time [62]. Furthermore, the Dynamic Bayesian Networks (DBNs) can simulate the seasonal variations of an area and provide vulnerability models that are characterized as more reliable and accurate [63]; thus, GALDIT takes into account uncertainty in climate projections, rather than relying solely on expert judgment. By transforming GALDIT from a static indicator to a climate indicator, future applications could better support sustainable groundwater extraction, early warning systems for extreme events, and dynamic management of coastal aquifers under changing climate pressures.

6.3. Data and Monitoring Improvements

Many researchers report and consider it necessary that updates should be made to the standard GALDIT vulnerability maps using GIS tools, in order to allow for continuous investigation and monitoring [40]. This can be done by combining new field data such as piezometric levels per season, new pumping test datasets for hydraulic conductivity, and recent water quality analyses (e.g., chloride concentrations and electrical conductivity). Each new dataset can be processed to create updated parameter layers (G, A, L, D, I, T), which are overlaid and reweighed to create an updated vulnerability index. Moreover, according to the same study, the AVI, GALDIT, and WQI indices are effective tools for synthesizing and presenting monitoring data to decision makers. Their application contributes to mitigating groundwater contamination resulting from seawater intrusion due to anthropogenic activities, while simultaneously offering a clear assessment of groundwater quality status [5].

6.4. Integration with New Tools

Moreover, to evaluate the increasing risks of groundwater salinity, further research is required. In general, several researchers have proposed machine learning tools to enhance the GALDIT index by incorporating additional parameters such as topography, land cover evolution, recharge variability, and pumping intensity, offering locally improved vulnerability estimates [64]. Although models such as GALDITMW have shown higher performance in complex environments by adding information on the porous medium and well density parameters [45], their application depends to a large extent on the availability and reliability of data. In addition, another option for future improvements is to emphasize time series analysis and data synthesis, allowing for periodic updating of vulnerability results based on monitored salinity trends. Also, GALDIT improvements with Machine Learning should not only be evaluated on the results obtained, but also on whether they effectively improve decision-making for groundwater management, where not only hydrogeological conditions but also ecological and environmental limitations must be accounted for. These improvements show that GALDIT is changing from a static vulnerability indicator to a flexible, dynamic, and integrated framework that is more capable of enabling the user to address the complex issues of coastal aquifer management.

7. Suggestions for Improving the Application SEAWAT Model

7.1. Geological and Hydrogeological Data Needs

The SEAWAT model can accurately simulate density-driven flow, but its results can be significantly limited by the availability of high-resolution geological and boundary data. In addition, in areas where coastal systems are laterally and vertically heterogeneous, fault geometry can introduce errors and uncertainties in predicted salinity fronts [34]. Tectonic discontinuities and fractured bedrock may form preferential pathways for seawater to intrude into coastal aquifers due to high-permeability, allowing density-driven flow to occur. This mechanism has been observed in areas where fractured aquifers promote seawater intrusion by increasing permeability. For larger coastal watersheds, this limitation can be reduced by coupling SEAWAT with watershed models, allowing for the dynamic transfer of precipitation, runoff, and recharge inputs to the groundwater model [65]. Additionally, the model runtime escalates when it incorporates small-scale heterogeneity, and more simplified models are often used to allow testing of scenarios on larger temporal and spatial scales.

7.2. Model Development and Calibration

Uncertainty in the parameters during model creation has been shown to affect the reliability of results, and further modeling with combined recharge values from climate change estimates has been suggested [66]. The incorporation of geologic, hydrochemical, and geophysical data into the model would render it increasingly able to discern between salinity sources, regardless of whether they are caused by upward migration of deep saline groundwater, evaporite dissolution, or anthropogenic inputs, and thereby support more targeted management and remediation activities. Although SEAWAT already couples MODFLOW with MT3DMS to simulate TDS transport, distinguishing between different salinity sources (e.g., deep brines vs. seawater intrusion) often requires additional supporting data, such as isotopic tracers, multi-depth sampling, and geophysical information, which can significantly improve model calibration and interpretation. Dunlop et al. [30] based their research on this issue, demonstrating the importance of TDS transport models, showing that deep brines can increase Cl concentrations and that the Simpson ratio, associated with the upward conical configuration, is a useful but unstable indicator. They also emphasized the need for supporting methods (e.g., isotopic analysis, multi-depth sampling, geoelectrical profiling) and automated calibration techniques to improve accuracy, particularly in regions where deep brines may be responsible for excessive chloride concentration [30,58,65]. Through the utilization of these supporting methods, firmer discrimination between salinity sources would be allowed, as well as enhanced accuracy for the model’s diagnostic and predictive capabilities.

7.3. Monitoring and Long-Term Data

A reliable simulation depends largely on the availability and reliability of the data, as well as on the temporal resolution of density-related parameters, especially chloride concentration, TDS, and hydraulic head. In many studies of coastal aquifers, the monitoring network is limited and provides only piezometric levels and/or chemical sampling, making it impossible to calibrate the evolution of the freshwater-saltwater interface. The lack of long-term and well-distributed data prevents reliable estimation of transfer, mixing zones, and boundary conditions, increasing the uncertainty in the long-term model scenarios [61]. Furthermore, the issue is not only the lack of data, but also the lack of appropriate temporal resolution, since information such as hourly or seasonal pumping records is often unavailable. Therefore, the data that ideally is needed are pumping rates, water salinity, and groundwater level seasonal fluctuations. In this way, SEAWAT can provide real and not theoretical predictions for salinization. SEAWAT predictive results are not only improved with more data but also by using specific time series on parameters such as chlorides (Cl⁻), TDS, pumping, and sea level changes. The lack of these data leads to models that do not reliably predict the future evolution of salinization [54]. Furthermore, the introduction of these measurements reduces uncertainty and makes SEAWAT models truly useful for management and adaptation to climate change conditions [67].

8. Similarities and Differences Between GALDIT and SEAWAT Models

GALDIT and SEAWAT certainly do not serve the same purpose, but can complement each other in an integrated decision-making framework. Furthermore, they are two important tools that are widely applied to coastal aquifers affected by seawater intrusion, but they serve fundamentally different roles. GALDIT is a parameter-based index designed to quickly identify areas that are potentially vulnerable to seawater intrusion under current hydrogeological conditions, especially when data availability is limited. The results of the GALDIT index do not provide quantitative results, i.e., how much the salinity front will increase after a number of years. In addition, GALDIT acts as a monitoring tool, providing a quick and low-cost delineation of high-vulnerability zones where aquifers are most vulnerable to seawater intrusion under current conditions. It provides a first assessment that helps researchers to determine where groundwater monitoring should be enhanced, and in which areas more frequent sampling should be performed. SEAWAT, on the other hand, is a physics-based numerical modeling code that simulates density-dependent flow and solute transport. It provides predictions of coastal aquifer salinization under different pumping conditions, sea level rise, and recharge variability. Because it requires detailed geological, hydraulic, and hydrochemical data, SEAWAT is most effective for the delineation of vulnerable zones. Therefore, while GALDIT supports spatial distribution of vulnerability and early warning, SEAWAT supports quantitative planning and long-term decision-making, showing that the two approaches are most effective when used sequentially rather than in isolation.

Table 1. A comparison of the main characteristics, advantages, drawbacks, and suggested applications of the SEAWAT numerical model compared to the GALDIT index-based approach for determining the susceptibility of coastal aquifers to seawater intrusion.

Feature	GALDIT	SEAWAT
Method Type	Index-based, parametric	Numerical, process-based
Purpose	Assess static vulnerability zones to seawater intrusion	Create a dynamic simulation of salinity transport and groundwater movement over time and space.
Key Parameters	Six hydrogeological parameters (G, A, L, D, I, T)	Full hydrogeological dataset: hydraulic conductivity, storativity, porosity, salinity, boundary conditions, etc.
Advantages	Low data requirements, quick and simple application, appropriate for areas with limited data, applied with GIS	High reliability, high realism, dynamic and predictive, and supportive of scenario analysis
Output	Static vulnerability map	Dynamic flow and salinity simulation
Limitations	Only offers subjective parameter weights and static assessments, it ignores changes over time.	Requires a large amount of high-quality input data, the calibration process is difficult and computationally demanding.
Recommended Use	Quick evaluation, first screening, and initial risk mapping	Comprehensive planning and management; assessment of control strategies, complex coastal aquifers
Best use	Screening and zoning	Detailed planning, scenarios
Benefit of Combination	Provides spatial zoning to guide targeted SEAWAT simulations	Validates and refines GALDIT maps with dynamic outputs
Typical Tools	GIS software (e.g., ArcGIS Pro 3.1., QGIS 3.34)	USGS SEAWAT, often coupled with MODFLOW and MT3DMS, various GUIs
Applications	Coastal areas, small unconfined aquifers	Large coastal systems with intensive pumping, climate change projections

summarizes their respective functions, data requirements, and recommended applications.

9. Discussion

GALDIT and SEAWAT are two of the most accepted methods utilized in the determination of coastal aquifer vulnerability to seawater intrusion, each with varying strengths and weaknesses.

GALDIT is a parametric, index-based model developed around six main parameters groundwater occurrence, aquifer hydraulic conductivity, depth to water table, distance from the shore, impact of current seawater intrusion, and thickness of the aquifer. It finds special utility in low-data availability regions due to its low cost, ease of use, and ease of rapid deployment. As an indicator-based method, GALDIT intentionally provides a static representation of vulnerability to seawater intrusion. This makes it a useful tool for rapid spatial assessment of vulnerability, but it cannot assess temporal evolution or management scenarios without being combined with dynamic models such as MODFLOW or SEAWAT. In response to this, several attempts have been made to extend or alter GALDIT by including additional parameters, site-specific weights, or by merging it with dynamic models.

SEAWAT, on the other hand, is a fully numerical model that is suitable for simulating density-dependent groundwater flow and solute transport, therefore especially suitable for simulating the complex hydrodynamic features of coastal aquifers. With the integration of MODFLOW (to simulate saturated flow) and MT3DMS (to simulate solute transport), SEAWAT allows scientists to conduct time-dependent simulations, test management scenarios, and model future climate or pumping regimes.

In the literature, there are few applications that combine the GALDIT method with SEAWAT simulations in coastal aquifers. The coupling of SEAWAT and GALDIT has thus become a practical hybrid strategy that allows users to balance the ease of use and wide range of applications of GALDIT with the process-based complexity and predictive strengths of SEAWAT. For example, Ref.[68] used these two GALDIT-MODFLOW models in combination to assess seawater intrusion into the coastal aquifer under different pumping and climate change scenarios. Their model was successful in identifying vulnerabilities and assessing the suitability of artificial recharge measures, validating the value of the contribution of static and dynamic modeling concepts. There is more innovation in the work of Ez-zaouy et al. [48], where the initial GALDIT framework was modified by incorporating three new parameters—hydraulic gradient, river contribution, and mapped extent of observed seawater intrusion—to enhance the incorporation of local hydrological and geological factors in the Souss–Massa basin of Morocco. In the Volturno River coastal aquifer (Italy), a semi-coupled HEC-RAS and 3D SEAWAT model demonstrated the ability to distinguish paleo-salinization from active seawater intrusion, emphasizing SEAWAT’s value in identifying the origin and timing of salinity sources. The research of Gaiolini et al. [69] is well-supported with regard to how modified SEAWAT models are able to simulate slowly developing, vulnerable, undetected salinization processes that are otherwise invisible to simpler models. According to the studies, the incorporation of such paleosalinity processes into coupled GALDIT-SEAWAT models is an innovative tool that can significantly enhance the applicability and robustness of coastal groundwater vulnerability models. In low-lying sedimentary coastal basins, especially those affected by past marine intrusions, land subsidence, or intense pumping, SEAWAT can provide a more realistic representation of salinity evolution by incorporating layered stratigraphy and variable density flow. These physics-based simulations can support and improve indicator-based assessments, such as GALDIT, by validating vulnerable zones and identifying hidden sources of salinity. Therefore, the future approach should focus on using GALDIT for rapid vulnerability profiling and SEAWAT for targeted validation and scenario testing, especially under the increasing impacts of climate change, sea level rise, and groundwater overexploitation.

10. Future Challenges

Significant progress has been made in hybridizing and modifying GALDIT and SEAWAT models to assess groundwater vulnerability to seawater intrusion. However, several research gaps and challenges remain and need to be addressed in future applications. Most calibrated GALDIT approaches for now remain semi-empirical and site-specific, limiting their generality in other hydrogeological and climatic conditions. Subsequent research needs to strive to develop universally applicable parameterizations, or regionally calibrated substitutes, supported by large-scale data. Moreover, while integration with numerical models like SEAWAT enhances process representation, it involves higher data needs and computational intensity that may be prohibitive in data-scarce or low-resource regions. Simplified or surrogate numerical models are required because they can maintain physical fidelity but reduce operation expenses. Future models should include climate and socio-economic factors like higher sea levels, extreme weather events, land subsidence, and human pressures like urbanization and irrigation expansion in order to make more accurate predictions. Remote sensing tools also provide a continuous and present image of the coastal environment (e.g., shoreline and land use changes, piezometric trends) while machine learning helps GALDIT–SEAWAT models to calibrate their parameters automatically from a large volume of data in order to increase the accuracy and reliability of their estimations under climatic and anthropogenic stress conditions. Their contribution is not limited only to data processing, but also allows one to assess the vulnerability and reduce the subjectivity in the weights of each parameter. Future developments should therefore focus on the use of machine learning-based calibration and remote sensing data to quantify uncertainty in order to produce dynamic models that allow more reliable management decisions depending on pumping, climate change or recharge scenarios.

11. Conclusions

Coastal groundwater aquifers are increasingly affected by seawater intrusion due to intensive groundwater abstraction, seasonal tourism demand, land use pressures and climate-related reductions in recharge. These stresses are intensifying groundwater salinization and highlight the need for vulnerability monitoring tools capable of predicting future intrusion under different management scenarios. The GALDIT method is also a tool that can be easily applied to identify areas susceptible to seawater intrusion, particularly in areas where hydrogeological data are limited. Several studies have improved the performance of the GALDIT index by using additional parameters, but these modifications are calibrated for the specific area and cannot be easily applied outside the study area.

Furthermore, SEAWAT provides a simulation based on the physics of density-dependent flow and salinity transport, which allows to find out how much water levels have decreased, the salinity front on a long-term scale, the sea level rise and what measures should be taken into account for the correct management of groundwater. Indeed, correct results of the model depend to a large extent on detailed geological, hydrochemical and temporal data, which are often difficult to obtain in coastal areas.

Thus, the two models serve complementary roles, i.e., GALDIT can be used as a first-stage screening tool to find zones of high and low vulnerability, while SEAWAT can then simulate the evolution of seawater intrusion. Furthermore, the results from SEAWAT simulations can be used to refine specific GALDIT parameters, such as groundwater level above sea level (L) and impact of existing seawater intrusion (I), improving the reliability of the vulnerability index. Future studies could use high-resolution monitoring data, such as isotopic analyses and geophysical measurements, to separate seawater intrusion from geogenic salinity. It can also be suggested that we should not depend only on hydrogeological data but should incorporate human and climatic factors into the studies. With improved datasets and computational developments, GALDIT-SEAWAT has strong potential to become an effective decision support tool for adaptive management of coastal groundwater.

Also, we propose a workflow for integrating GALDIT with SEAWAT, which is summarized in Figure 6. First, GALDIT is applied to classify the coastal aquifer into low, moderate, and high vulnerability zones. Only the moderate and high vulnerability areas are selected for further analysis. SEAWAT input data are then prepared, including hydraulic properties, pumping rates, salinity observations, recharge values, and local stratigraphy. Moreover, the density-dependent flow simulations are performed using SEAWAT under different scenarios. Finally, SEAWAT results are used to identify GALDIT parameters, particularly groundwater level above sea level (L) and the impact of existing intrusion (I), allowing an updated and more reliable static vulnerability assessment. This structured workflow demonstrates a practical approach for coupling the two models and ensures methodological transparency.