Evaluating the Effectiveness of an Interactive Tool for Water Governance in Transboundary Basins: A Participation-Based Approach and Visualization of Water Security from a Vulnerability Perspective

Abstract

In this study, we analyzed the effectiveness of a data visualization tool (dashboard) designed to provide insights to decision-makers about the vulnerability of water resources in the tri-national Upper Lempa River Basin (CARL) in the face of future climate and land use uncertainties. The effectiveness of the dashboard was assessed using three methods: (1) a user survey for evaluating dashboard clarity, completeness, and ease of use using seven parameters proposed by Stephen Few. The result of the survey overwhelmingly indicated a positive experience when interacting with the dashboard; (2) pre- and post-use tests were used to assess knowledge acquisition. The users’ correct answers increased by an average of 35%, and incorrect answers decreased by an average of 25% for questions assessing an understanding of water demands, the effects of climate change on the basin’s water security, land use trends, and the sub-basins with the highest vulnerability in the region; and, significantly, (3) user reports on insights drawn from their interaction with the dashboard. Users drew insights on the extent to which different regions will likely experience increased vulnerability regarding water resources and on strategies that could reduce this vulnerability. The key issues that need to be addressed to ensure that the dashboard fosters greater transparency and public participation in governance and is used by decision-makers to improve the management of water resources in this watershed are the following: (1) the incorporation of user feedback and the inclusion of adaptation strategies and their impacts on the dashboard; (2) the commitment and training of a local institution to host and maintain the dashboard and to make it available to the public; (3) the engagement of decision-makers from the three countries regarding the use of the dashboard to understand future uncertainties and the potential impact of adaptation strategies on performance metrics.

1. Introduction

Water governance can be defined as the set of processes, institutions, legislation, and policies that drive decision-making around water resource management and development through the services it provides to society [1,2]. Globally, pressures on water resources are increasing due to the effects of factors such as climate change, population growth, changes in land use, and anthropogenic activity on ecosystems [3]. Consequently, effective water governance is fundamental for water security and environmentally sustainable, socially sustainable, and equitable development [4]. Achieving functional water governance in a transboundary context is more challenging since international river basins with shared political boundaries imply the involvement of different nations with diverse policies, interests, and priorities and, thus, the need for greater coordination within water governance.

The approach to water governance from a transboundary perspective requires a particular focus on ensuring the participation of all stakeholders within the basin context to build trust and a spirit of cooperation among them [5]. In this way, there will be greater cooperation between the countries for the establishment of agreements that support the joint management of water resources; this considers the different levels of management of the stakeholders involved, including local governments, communities, non-governmental organizations, and the private sector, ensuring the transparency of decision-making processes [6]. Among the many issues addressed in this decision-making process is water security. The United Nations defines water security as the capacity of a population to safeguard sustainable access to adequate quantities of acceptable quality water for sustaining livelihoods, human well-being, and socioeconomic development, ensuring protection against water-borne pollution and water-related disasters and preserving ecosystems in a climate of peace and political stability [7]. Thus, water security highlights the importance of an interdisciplinary and multidimensional approach to its development.

Interactive tools such as dashboards for data visualization emerge as useful instruments to facilitate decision-making under a water security framework [8]. This type of tool allows the representation of large volumes of information in a more understandable and accessible way, also considering the integration of multiple variables and the analysis of potential future scenarios. This provides a better understanding of the different problems related to water resources, promotes a more active participation of the different stakeholders involved, and increases the understanding of the impacts of the decisions to be made within water governance [9].

Previous research on the development of dashboards for water management has provided valuable insights, highlighting both achievements and persistent challenges in this field. The gaps that this research seeks to address arise from these challenges. The research in [10] presents the design of a dashboard of 52 indicators for a case study of Pakistan and a regional comparative analysis between countries. When evaluating the efficiency of the dashboard, they observe that it facilitates a rapid assessment of water security, identifying trends and challenges, which increases vulnerability to water risks. In the process of evaluating the use of dashboards in urban water security in [9], it is observed that the main factors in which their effectiveness depends on are the visual design and user customization. In particular, the integration and combination of these two factors is essential to promote the development of an environmentally effective dashboard.

The contributions presented in [11] highlight that the use of dashboards in water treatment plants helps to improve factors such as speed and accuracy in operational decision-making. It is highlighted that the dashboard generates important contributions through processes of data combination, the integration of previous experiences, and scientific prioritization, which are essential in order to find optimal solutions quickly, positively impacting public health, the environment, and operational efficiency. According to [12], the implementation of an intelligent dashboard integrated with automatic water meter reading devices contributes to reducing operating costs and improving the quality of service in water companies. The result of the implemented tests enabled to recognition of the effectiveness of this technology in reducing non-revenue water (NRW) and in cost savings. Similar findings presented in [13] demonstrate that implementing an integrated water management system in Italy, using dashboards, significantly improved operational efficiency and business performance in the water industry. Digitalization facilitated agile responses to operational needs and improved business decision-making, favorably influencing water management.

On the other hand, as highlighted in [8], dashboards can enhance water resource management by enabling the exploration and analysis of complex data in a logical and simultaneous manner, thanks to advanced user interaction techniques, data transformation, and coordinated views. These findings align with those presented in [14], where dashboards are described as intelligent water management systems, with flexibility and integration capacity. They facilitate the acquisition and visualization of data in water distribution networks, overcoming limitations of commercial platforms and increasing interoperability. Finally, in [15], aspects such as the flexibility and integration capacity of dashboards are highlighted, facilitating the acquisition and visualization of data in water distribution networks, increasing interoperability, an essential factor when referring to the case of transboundary basins.

The analysis that has been proposed shows that the authors have been highlighting the importance and value of dashboards as tools that help to significantly improve water resource management, especially when there are processes of cross-border dependency, as dashboards promotes collaborative approaches and effective data processing. These benefits are even more evident when dashboards are integrated into smart management systems or are optimized using various types of technologies connected to the Internet. However, the lack of studies that point out, from a critical perspective, deficiencies in the design and use of dashboards in water management is evident, which is essential to promote the continuous improvement of these types of tools. There is also a lack of studies that evaluate the effectiveness of dashboards through studies focused on the analysis of experiences and perceptions of people for whom these tools are designed, especially in spaces that promote the collective construction of knowledge around water management.

Under these considerations, and taking into account the water security problems experienced by the transboundary Upper Lempa River Basin (CARL) due to changes in land use, variations in rainfall and evapotranspiration regimes, and changes in water demand, among others, an interactive dashboard was designed to address the analysis of the vulnerability of the CARL, as it encompasses the territories of Guatemala, El Salvador, and Honduras. We introduce uncertainty under the Robust Decision Support (RDS) methodology, which explores different future scenarios to assess the vulnerabilities of a watershed [16]. In the RDS approach, uncertainty is understood as a series of factors and future conditions that can affect decision-making in water resources management, such as climate change, variations in water demands, socioeconomic and demographic changes, and government policies, among others.

When designing such dashboards, it is common for researchers to assume that these tools adequately meet the objectives for which they were developed. However, this assumption may overlook the actual experiences of users interacting with them, which creates a gap in the understanding of their effectiveness and usability [17]. Therefore, this study presents the process of developing the CARL vulnerability dashboard and its performance evaluation within the water governance process that has been implemented in the CARL project to address water security in the region. This project is the result of a cooperative agreement by the Sustainable Water Partnership (SWP) funded by the United States Agency for International Development (USAID) and implemented locally by WINROCK International; the objective is to address the water security risks associated with climate change and anthropogenic activities in the CARL with practical solutions to improve climate resilience and governance regarding the region’s natural resources [18].

Finally, this study evidences its contribution to strengthening knowledge on the vulnerability of water resources, especially on the impact of land use change and the identification of critical regions. The tool facilitated the understanding of complex data and its translation into practical information for decision-making, highlighting the importance of a clear and accessible interface. It also promoted greater participation and the co-responsibility of local stakeholders, fostering a collaborative and transparent approach to transboundary water governance. These contributions highlight the technical, social, and participatory value of the dashboard in informed decision-making.

2. Materials and Methods

2.1. Study Area

The Lempa River basin is a tri-national basin with an area of nearly 18,000 km², covering three Central American countries: El Salvador, Guatemala, and Honduras [19]. The Lempa River is in this basin, which is the second longest river in Central America and flows into the Pacific Ocean; it is also one of the most important rivers for water supply in El Salvador [20]. The upper basin of this watershed is the subject of study in this research, with an approximate area of 4343 km² (distributed in 33 municipalities), where 58.6% is Guatemalan territory (17 municipalities), 29.5% is El Salvador territory (11 districts), and 11.6% is Honduran territory (5 municipalities) [21]. It has two main drainage areas known as “Río Lempa Alto” (passing through Guatemalan and Honduran territory), with a channel length of 107.7 km, and “Angue-Ostúa-Güija” (mainly located in Guatemala until it flows into Lake Guija), with a channel length of 211 km, both flowing into the Desagüe River in El Salvador [22].

The most important water bodies in the basin include Lake Güija, shared by Guatemala and El Salvador, Laguna Metapán in Salvadoran territory, and Laguna Atescatempa in Guatemala. This basin is distributed in seven main sub-basins: the Ostúa river, Angue river, Guajoyo river, Cuspama river (Atescatempa Lagoon), Güija lake, Metapán lagoon, and the Upper Lempa river (see Figure 1).

According to [23], the CARL faces a significant problem due to the growing demand for water and the decrease in its supply because of various factors that have altered water behavior over the last 30 years. These factors include changes in land use, variations in rainfall and evapotranspiration regimes, and changes in water demand, which have led to a reduction in flow of up to 70% in the dry season. These problems are compounded by the pollution of water resources due to poor solid waste management, untreated wastewater, and agricultural practices that pollute the water with agrochemicals. These factors, in addition to deforestation due to changes in land use and agricultural and livestock activities, have deteriorated water quality, seriously affecting the well-being of the population.

To address these problems associated with water security, the dashboard presented in this research was developed. This tool makes it possible to identify water resource vulnerabilities and identify the most affected areas, facilitating the analysis of risk factors. In addition, the dashboard provides a basis to guide the process of identifying adaptation strategies and measures, contributing to a more efficient and sustainable management of water resources in the region.

2.2. Dashboard Design and Development

The design process of the interactive dashboard for the vulnerability analysis of the CARL was structured in three main phases: (1) understanding the current and future conditions of the basin under uncertainty; (2) characterizing vulnerability using water security indexes; and (3) integrating data visualization in interactive dashboards. In the first phase, the current and potential future scenarios were characterized and integrated into the water resource collaborative model using Water Evaluation And Planning (WEAP Version 2024.0006 https://www.weap21.org/ USA) [24], with the aim of evaluating their impact on the biophysical and demand variables of the CARL. The second phase focused on the approach of methods for estimating vulnerability by using different indexes related to water security in terms of water quantity and quality, and, in addition, the use of the concept of failure frequency [25] for estimating the magnitude of vulnerability over time according to minimum acceptable values for these indexes (the information used for this phase also comes from the model developed in WEAP). Finally, in the third phase, we worked on everything related to the graphic representation of vulnerability through an interactive dashboard in Tableau (Version 2023.1 USA) [26]; this would allow users to explore the results in a more accessible and understandable way, considering their preferences and thus facilitating informed decision-making within a context of water governance (all considering scenarios of climate change and land use change). The three phases are explained in detail below.

It is important to note that the information used in the development of the water resources model, which served as input for the creation of the dashboard, comes mainly from official sources provided by institutional and governmental entities. To a lesser extent, data from satellite repositories were used when there was no information available for a specific variable, thus complementing the database used for the analysis.

2.2.1. Understanding the Uncertainty in the Basin

Uncertainty is a topic that has been studied in different areas, such as natural, social, and management sciences. In fact, its role has gained importance in research on environmental changes, including climate change [27]. According to [28], uncertainty is associated with potential future conditions that generate variability and may impact the performance of water resources. In the RDS approach, uncertainty is understood as situations in which the actors in a decision lack clear agreement or knowledge about the most appropriate model for linking actions to their outcomes or for estimating the probability of future events [29]. Thus, uncertainty in this research refers to future scenarios that reflect different possible conditions in the future.

Consequently, the incorporation of uncertainty is fundamental for this research to achieve a more robust understanding of the different ways in which a system (in this case, the CARL) can respond to different future conditions. Two sources of uncertainty were evaluated in the analysis: The first source is associated with global changes such as climate change [30] and the second source is land use change [31], particularly agricultural use. This was implemented through the development of scenarios that simulated potential future trajectories, as well as their impact on water resources in both the short (2021–2040) and medium term (2041–2060). The following is a brief description of the methods applied to incorporate uncertainty.

• Uncertainty associated with climate change:

Future climate conditions are a relevant uncertainty factor for water resource management; moreover, how climate will evolve and how it will affect future water availability are important unknowns to consider. To analyze these uncertainties, this work considered the climate projections presented in the Sixth Assessment Report of the Intergovernmental Panel on Climate Change [32]. The WEAP model was created using the climate information of 20 General Circulation Models (GCM) that were evaluated in the four Tier-1 Shared Socioeconomic Pathways scenarios (SSP scenarios 1–2.6, 2–4.5, 3–7.0, and 5–8.5). The selection of GCM was based on both the availability of information needed for the WEAP model (precipitation, mean temperature, and monthly temperature extremes) and the fact that they have been evaluated in the four SSP scenarios. The climate information and results of the WEAP model were grouped to represent dry, wet, and trend climates.

• Uncertainty associated with land use change:

Land use changes in agriculture are inherently linked to the socioeconomic dynamics of the populations living in the CARL. They are also related to environmental factors such as wildfires and droughts, which were identified in fieldwork implemented during the study. To represent the uncertainty of the land use change in the CARL, a mapping of the behavior of the main crops in the CARL during the last 20 years was carried out. This made it possible to identify relative stability in the patterns of change in agricultural activities, implying that the agricultural frontier has not grown significantly. Most of the agricultural area is concentrated in six types of crops (tomato, agroforestry plantation, coffee, grains, vegetables, and melon).

We used the results of the analysis of the agricultural area in the CARL developed in [33], where the authors found the tendency of changes in the six types of crops using the Normalized Difference Vegetation Index (NDVI) [34] and the Man–Kendall test [35]. These identified trends were used to project changes over the 40-year period (projected). Figure 2 shows the projected change for six crops; the results suggest an increase in the cultivation of basic grains (corn and beans), vegetables, and tomato, with a homogeneous variation in agroforestry plantation and melon and a decrease in coffee crops. These projected trends have been incorporated into the vulnerability analysis under two scenarios: one with trend changes, as shown in Figure 3, and one with no changes, which assumes that land use in the future remains the same as in the base period between 2000 and 2019.

2.2.2. Vulnerability Estimation

A growing challenge facing Integrated Water Resource Management in a basin context is water security due to the confluence of various factors such as climate change, climate variability, population growth, changes in land use, and the contamination of water sources, among others [36]. Water security refers to the ability to ensure sustainable access to sufficient and quality water, protecting livelihoods, well-being, socioeconomic development, and ecosystems in a stable and uncontaminated environment [7]. All these factors have a negative impact on the available water resources both from a quantity and quality perspective, which simultaneously implies affectations on ecosystem health, drinking water supply, energy, food security [37], and other water uses for different sectors. Particularly, in the CARL, a significant increase in water demand has been experienced during the last three decades because of population growth and agricultural activity. In addition, the water supply has decreased due to changes in land use and alterations in the rainfall regime, causing a reduction of up to 70% in flow during the dry season. The basin also has serious water resource contamination problems due to untreated wastewater discharges, improper solid waste management, and deforestation caused by the expansion of the agricultural and livestock frontier [21].

To estimate the vulnerability of the CARL to these problems, five indexes associated with water quantity and quality were used: the Standardized Precipitation Index (SPI) to evaluate the variability and availability of water through changes in precipitation for 3 and 6 months [38]; the Water Use Index (WUI) calculated the relationship between water demand and supply [39]; the Groundwater Regulation Index (GRI), as an indicator of groundwater recharge regarding aquifers; the Water Quality Index (WQI) to estimate the concentration of water pollutants from physical, chemical, and biological conditions of water, and the Potential Alteration Index of Water Quality (PAIWQ), which focuses on the evaluation of pollutant loads in the discharge of contaminated water into water sources [39]. Finally, all these indexes were weighted in a global Water Security Index (WSI) with the objective of reflecting the state of water security in general terms (combining aspects related to both water quantity and quality).

The selection of the five water security indices used to estimate the hydric vulnerability of the CARL is based on their ability to capture key dimensions of water security in a comprehensive and multidimensional manner. Each index addresses a critical aspect of water resource dynamics, ensuring that both quantity and quality components are adequately represented. The SPI evaluated the variability and availability of water through changes in precipitation, crucial for understanding droughts and water surpluses [38]. The WUI measures the balance between water demand and supply, highlighting pressures on the resource [39]. The GRI assesses the recharge and depletion dynamics of aquifers, a key factor for long-term sustainability. The ICA and the PAIWQ provide insights into the state of water quality and the impact of pollutant loads, respectively [39]. Together, these indices offer a holistic view of the basin’s water vulnerability, addressing both physical and anthropogenic drivers of risk. Their integration ensures a robust assessment framework that supports informed decision-making, and the design of effective adaptation strategies tailored to the basin’s specific challenges.

The GRI is an index proposed by this research and represents the level of recharge or depletion of a groundwater system with respect to the local and historical conditions of the basin. Positive values indicate aquifer recharge, while negative values indicate aquifer depletion without considering withdrawals. It is calculated based on the net recharge (R) of each aquifer, standardizing it jointly by considering all the aquifer systems of the basin by means of a net recharge matrix, where the rows and columns represent each time instant and each aquifer sector, respectively, as follows:

$$\mathrm{G}\mathrm{R}\mathrm{I}=0.5 \mathrm{\ast } \left(1+{\displaystyle \frac{\mathrm{R}}{\mathrm{m}\mathrm{a}\mathrm{x} \left(\mathrm{a}\mathrm{b}\mathrm{s}\right(\mathrm{R}\left)\right)}}\right);\hspace{1em}\mathrm{R}=\mathrm{I}-{\mathrm{F}}_{\mathrm{b}}$$

(1)

where I is the infiltration to the subway system, and F_b is the baseflow of each aquifer system.

2.2.3. Data Visualization

A dashboard is a visual tool that displays data to monitor situations and/or improve understanding, using visualizations that respond to a narrative [40]. The narrative proposed for this research arises from what knowledge is required to understand water security from a context of vulnerability and, thus, be able to improve it. According to [41], these requirements are divided into four categories: (a) information on the historical context of water in the region and the perception of water problems and their solutions according to that context; (b) the exogenous factors that determine the quality and quantity of available water; (c) the endogenous tools to influence the balance of water resources; and (d) the impacts of strategies to address the main problems. Later, the authors of [10] added three more requirements for the water security framework: (e) a water resource system, (f) system performance indicators, and (g) trends, identifying the need to describe the variables of the water system; these are used to assess how the water system is operating and to report on the evolution of water security in the face of possible future scenarios, respectively.

Based on this, a narrative order was established in the dashboard to frame the water security of the CARL. This order is composed of three main components: (1) Context, (2) Indexes (of water security), and (3) Vulnerability Analysis (see Figure 3). The Context component is related to Briscoe’s requirements (a) and (b) and the additional (e); the Indexes component is related to the requirement (f); and the Vulnerability Analysis component is related to factor (g). In summary, this narrative focuses first on understanding the biophysical system context (both historical and potential future), then on indicators as an endogenous tool for identifying vulnerability trends influenced by exogenous factors; finally, this vulnerability analysis is complemented by an approach that considers not only the extent of vulnerability but also its frequency.

The Context component (first component) provides basic information on the methods and evaluation of the biophysical variables simulated in WEAP (climate, hydrology, demands, and water quality) to gain a basis for understanding the situation of the watershed under different scenarios. In the second component, Indexes, the results of the water security indexes are represented (associated with water security and supply) to identify the principal vulnerabilities of the basin in terms of water quantity and quality and in the face of different scenarios. Finally, the Vulnerability Analysis component incorporates the visualization of the frequency of failure for the water security indexes and supply metrics; this enables an approximation of the portion of time in which water needs cannot be met (controlled using minimum acceptable thresholds for the indexes which vary according to the control panel) under different scenarios (see Figure 3).

The dashboard has two visualization types at a temporal scale: annual and monthly, except for the water security indexes, which were estimated only on an annual scale (see Figure 4). First, in the annual scale visualization, two types of graphs are used: a map graph and a time series graph (see Figure 4a). The map makes it possible to contrast the historical condition with the projected conditions for potential future scenarios (both climate change and land use change). This contrast is presented using a single average value of the variable associated with the selected climate condition for each sub-basin of the CARL, allowing for the selection of the analysis period between the short and medium term. The second type of graph corresponds to a time series that facilitates comparing the historical condition and the range of future possibilities of the analyzed variable for each sub-basin. This range of possibilities was represented by using different percentiles associated with extreme conditions: P05 for extremely low values represented by 5%, P50 for mid-point values represented by 50%, and P95 for extreme high values represented by 95%. The map-type graph was used to provide users with a spatial reference of the behavior of each variable at the sub-basin level, allowing for a detailed geographic visualization. On the other hand, the time series graph was used to analyze the evolution of the variable over time. Additionally, the previously described percentiles were incorporated, with the objective of offering a comprehensive view that encompasses the totality of possible future values of the variable under study based on the climate change simulations carried out.

The second type of time scale visualization corresponds to the monthly scale, for which two types of graphs are available (see Figure 4b). The first is a map like the one used in the annual visualization but with the possibility of selecting a specific month to visualize the single average value associated with the sub-basin. The second is a boxplot graph, the objective of which is to show the distribution of the data at the monthly multiannual level for the analyzed variable, both at the sub-basin level and for the basin. This graph allows for the comparison of different future conditions with historical conditions (purple line), providing a clear representation of the magnitude of the average change in the variable from month to month in relation to its historical behavior.

On the other hand, the dashboard allows for exploring the uncertainty of future projections by using the two types of scenarios explained in Section 2.2.1. The first type of uncertainty related to climate change is the “Climate Condition” scenario, which enables the navigation through the climate change scenarios grouped into “Dry Climate”, “Trend Climate”, and “Wet Climate”. These scenarios can be explored both spatially (maps) and temporally (multi-year monthly boxplots). The second type of scenario is related to land use change, called “Soil scenarios”, where a “No change” scenario can be explored when the observed soil conditions are maintained, and a “Trend” scenario, where the crop change trends explained in Section 2.2.1. are projected.

The following is a brief description of each of the dashboard components.

• Context component

The Context component consists of five different types of windows: (1) Climate, (2) Hydrology, (3) Aquifers, (4) Demand, and (5) Water Quality. In general, the objective of these windows is to display information related to the balance between water demand and supply in terms of quantity and quality at the level of the CARL and its sub-basins. It also provides an information base for understanding the security indexes since these indexes are directly related to the biophysical variables.

Users can explore three critical variables in the Climate section: precipitation (mm), temperature (°C), and evapotranspiration. In the Hydrology section, information is provided on the flow at the outlet of the sub-basins (m³/s), the surface supply (m³), which includes the sum of baseflow, interflow, and runoff, and the groundwater supply (m³), which represents the difference between all contributions to the aquifers from the surface system and those that the groundwater system returns to the surface system (baseflow) without considering extractions. The Aquifers section complements the Hydrology section by showing the spatial distribution of aquifers in the basin. In the Demand section, a map shows the spatial distribution of agriculture and population centers, disaggregated into rural and urban areas, together with the water demands for agricultural and domestic use (rural and urban). Finally, the Water Quality section presents a map with the geolocation of water quality-monitoring points, as well as information on pollutants in surface water, both in concentration (mg/l) and load (kg), including BOD, COD, fecal coliform (FC), total phosphorus (P), total nitrogen (NO₃), and total suspended solids (TSS).

• Indexes component

The water security indexes are displayed in the second component of the dashboard, initially accompanied by a concise description of the index concept, as well as a summary of the estimation method. The description details the values and interpretation ranges of the indexes used to assess water quantity (SPI, WUI, and GRI) and water quality (WQI and PAIWQ). Although the indexes present a variety of ranges in their conventional reporting values, in the vulnerability analysis, these ranges have been normalized between 0 and 1, as explained in Section 2.2.2. This component of the study is structured in two windows: (1) WSI (biophysical) and (2) Supply Metrics; the latter is a complement to the understanding of vulnerability from the supply capacity to the population centers, with variables such as urban and rural coverage (%) and urban and rural per capita supply (L/hab*day).

• Vulnerability Analysis component

By transferring the IPCC definition of vulnerability [42] to a water security context, vulnerability refers to the susceptibility of a water system to alterations under different stress conditions. This is evaluated based on three fundamental dimensions: exposure, sensitivity, and adaptive capacity [43]. First, exposure corresponds to the level at which a water system is subject to threats that can alter the quantity and quality of available water, such as climate change or land use change. Secondly, sensitivity is related to the measurement of the response of this system through the evaluation of the impacts that these threats may have on the water resource. Finally, adaptive capacity is associated with the ability of the system to adjust to hazards effectively.

Through the implemented analysis, vulnerability was approached from its exposure dimension. A modified version of the methodology described in [25] was implemented, which is structured as a three-step process: (1) a time series analysis of the WSI and other metrics for vulnerability estimation, (2) vulnerability calculation, and (3) decision space visualization. A brief description of each of these steps follows:

• Time series analysis of the WSI and other metrics: This corresponds to the temporal visualization of the modeled results from the XLRM (X: uncertain eXogenus factors, L: Levers for decision-makers, R: analytical tools to Relate uncertainties with levers, and M: Metrics of performance to evaluated outcomes) approach of the index under different uncertainties associated with vulnerability. The objective is to identify desirable performance levels that reduce the basin’s vulnerabilities through management strategies or policies. This visualization was implemented in the index component described above.
• Vulnerability calculation using conversion to a single-value form: In this step, the temporal data of the WSI are consolidated into a single value according to the performance of the system with respect to desirable levels defined by the stakeholders. In the case study, vulnerability was calculated as the equivalent of the Failure Frequency (from the reliability calculation method of [44]) of the system, with this being the percentage of times that the WSI falls below the desired threshold under different uncertainty conditions:

$$F_f={\displaystyle \frac{{\sum }_{t=1}^T{E}_t}{T}}$$

(2)

where T is the total number of model time steps, and t is each individual time step; $E_t=1 if {IndexWSI}_t<Threshold or E_t=0 if {IndexWSI}_t>Threshold$. We worked in this way since the lower values of the different WSIes represent greater vulnerability.

3.

Visualization of the decision space: This graphical representation consists of four main components: (a) a menu for the selection of the index to be analyzed and the different scenario options; (b) a map showing the spatial display of the failure frequency of the selected index at the sub-basin level; (c) a control panel to adjust the thresholds corresponding to the different indexes; and (d) a graph in grid format presenting the failure frequency matrix of the selected index.

Selection menu: This menu allows for the selection of the specific index to be analyzed, as well as the climate change condition (dry, trend, or wet), the soil scenario (no change or trend), and the time horizon (short term or medium term).

Failure frequency map: The map facilitates a comparison between the historical condition and the projections of possible future scenarios (according to the selection made) using an average value for the frequency of failure for each sub-basin.

Threshold control panel: The panel includes a slider for each of the water safety indexes, allowing a value between 0 and 1 to be assigned to evaluate the impact of different threshold values on the failure frequency displayed in the matrix. This threshold defines the minimum value that the index must reach, from which the failure frequency is calculated. This functionality facilitates the vulnerability assessment by allowing for the immediate visualization of the impact of the different thresholds selected by the user for each index, thus improving the process of informed decision-making.

Failure frequency matrix: This matrix shows the results of the previous step, where a higher vulnerability (values close to 100%) is indicated with colors close to red and a lower vulnerability with colors close to green (values close to 0%). In the rows of this matrix are the historical condition and potential future scenarios, and in the columns are the sub-basins of the CARL. In the dashboard, this matrix is interactive (see Figure 5); it changes according to the selection of index, scenario, period, and threshold value. This approach facilitates an understanding of the possibility of system failure at the desired threshold, allowing for instantaneous temporal and spatial contrasts for the preliminary identification of management strategies to address this vulnerability so that, in the next phase, the vulnerability of the basin can be analyzed from its sensitivity dimension with the evaluation of the impact of such strategies.

2.3. Effectiveness Assessment Approach

To determine whether tools such as the CARL vulnerability dashboard contribute to strengthening water governance, a methodology was proposed for its evaluation from two approaches. The first is related to capacity building regarding the vulnerability of the basin related to the stakeholders who were involved in the vulnerability workshop implemented in El Salvador as part of the RDS approach. The second is associated with the interaction of the stakeholders with the dashboard during the participatory decision-making process and their perception of the effectiveness of this tool.

The implementation of a pre- and post-use knowledge test regarding this interactive visualization tool is a vital part of the objective evaluation of its impact on the capacity building of stakeholders concerning the topic under analysis, which, in this case, is the water vulnerability of the CARL. Thus, it is possible to contrast the initial knowledge of the stakeholders of the issue and the knowledge acquired after interaction with the tool [45]. This allows us to determine how effective the tool is in transmitting key concepts and strengthening stakeholders’ capacities to improve the decision-making process, considering that this is part of the objective for which the tool was designed.

The knowledge test was structured according to the following questions associated with understanding the water vulnerability of the CARL:

• What factors do you consider to contribute to the vulnerability of the CARL in terms of water quantity and quality?
• What are the main water uses in the CARL, and how might they relate to the vulnerability of the watershed?
• How do you think climate change could affect water security in the CARL?
• What are the trends in land use change in the CARL, and how might these trends affect the quantity and quality of water available in the basin?
• Which regions do you think are the most vulnerable in the CARL, and for what reason?

The tool’s effectiveness was also evaluated based on the stakeholder’s perception during their interaction with the dashboard, focusing on the ability of the visualizations presented to communicate effectively and provide an adequate understanding of the topic under analysis. Since this understanding can be approached from different perspectives, seven criteria were used to measure such effectiveness. For this purpose, the data visualization effectiveness profile proposed in [46] was implemented, which establishes five criteria under an informative approach (focused on generating knowledge using visualizations: usefulness, completeness, perceptibility, truthfulness, and intuitiveness) and two criteria under an emotional approach (focused on developing a constructive emotional response: esthetics and engagement/interest). Based on these criteria, questions were designed and applied in a participatory space with watershed stakeholders.

Table 1. Criteria for evaluating stakeholder interaction with the dashboard (adapted from [46]).

Criteria	Description
Usefulness	This refers to the ability of the tool to transmit information that is useful to the user, i.e., that the tool communicates information that is of value to the user in their decision-making process. The value of the information is a subjective aspect, as it depends on the user’s needs.
Completeness	This refers to the ability of the tool to produce a desired level of understanding by presenting the right information and in the right amount, not presenting more information than is required to achieve the knowledge required by the user. Additionally, it concerns the ability to provide the necessary context to understand the information.
Perceptibility	This refers to the ability of the tool to display information that the human eye and brain can perceive with minimum effort and adequate accuracy.
Truthfulness	This refers to the degree of accuracy and validity of the information presented through the tool. Accuracy is a measure of reliability and precision. Validity indicates how well the phenomenon to be represented is being represented.
Intuitiveness	This refers to the tool’s ability to be easily understood and usable, which feels familiar to the user.
Esthetics	This refers to the ability of the tool to be pleasing to the human eye, considering that the user is likely to receive the information presented to them better if the tool is “beautiful” or esthetically pleasing.
Engagement/Interest	This refers to the ability of the tool to invite the user to examine the information and generate interest in using it to gain knowledge of the phenomenon to be represented.

details the criteria corresponding to each of these categories, as well as the associated questions and their possible rating according to the Likert scale [47].

The knowledge tests, administered before and after the use of the dashboard, as well as the survey related to user interaction during the vulnerability workshop, were conducted with a group of 20 participants from national government institutions, local governments, non-governmental organizations, academic institutions, and regional cooperation bodies. Ethical protocols were followed for the collection, use, and sharing of the information obtained, ensuring informed consent from participants and safeguarding data confidentiality. The information was collected anonymously, without recording participant’s identities, to protect their privacy and foster a transparent and unbiased environment for participation.

Additionally, for the design of the dashboard and the participatory exercises implemented in the vulnerability workshop, the guidelines for data visualization effectiveness established in [48] were considered, as well as the Information Search Mantra from [49,50]. This protocol establishes a series of key steps to optimize interaction with complex data visualizations. Five of the seven actions considered within this protocol are described below:

• Overview: This refers to a broad view of the entire visualization, understanding the big picture before focusing on details. It should also contain a moving view box to control the content of the view.
• Zoom: To drill down on specific elements, the user should have the option to zoom in on sections of interest, using tools that allow them to adjust the focus and zoom factor according to their needs.
• Filter: This is an option for filtering data to focus on relevant items, allowing users to sharpen their focus and view only information relevant to their specific interest or questions.
• Detail: Users can access more detailed information by selecting an item or group of interest, facilitating a deeper understanding of the selected data.
• Relate: This refers to options in the visualization for users to identify relationships between elements within the visualization, using dashboard capabilities to link variables and explore possible causalities or connections between phenomena.

Figure 6 illustrates the flow diagram of the methodology implemented in this research. It shows the phases prior to the development of the dashboard: the modeling of water resources, the characterization of uncertainty, and the estimation of vulnerability; the phase of data visualization in the dashboard; and the phase after the development of the dashboard—the evaluation of the effectiveness of the dashboard.

3. Results

3.1. Impact of the Dashboard on the Understanding of Water Vulnerability

To determine whether tools such as the CARL vulnerability dashboard contribute to strengthening water governance, a methodology was proposed for its evaluation using two approaches. The first is related to capacity building regarding the vulnerability of the basin related to the stakeholders involved in the vulnerability workshop implemented in El Salvador as part of the RDS approach. The second is associated with the interaction of the stakeholders with the dashboard during the participatory decision-making process and their perception of the effectiveness of this tool.

Figure 7 presents the change when comparing the pre- and post-dashboard tests, identifying the impact on the capacity building of the actors with respect to the issue addressed by each question; the following list details the meanings of the possible response categories:

• Incorrect: The answers do not reflect an adequate understanding of the concepts.
• Partially incorrect: The answers reflect a partial understanding of the concepts. According to the topics addressed using the dashboard, part of the answer is correct, but part is not.
• Correct: The answers are aligned with the key concepts regarding the water vulnerability ability of the watershed.

Question 1, associated with the factors affecting the vulnerability of water resources in relation to quantity and quality, showed about a 10% drop in the number of incorrect answers. Nevertheless, there has been no growth in correct answers, which indicates that the dashboard has facilitated only a partial improvement in the understanding of these factors. This result indicates the need to reinforce key concepts related to the multiple factors of vulnerability of water resources.

On the other hand, Question 2, which explores water users in the basin and their relationship to vulnerability, showed substantial improvement, with a 25% decrease in incorrect answers and a 50% increase in correct answers. These results suggest that the dashboard was particularly effective in clarifying the relationship between different water uses and their impact on basin vulnerability. The tool allowed participants to better understand how current and future water uses can amplify or mitigate the risks associated with the vulnerability of water resources.

In Question 3, regarding the effects of climate change on the basin’s water security, while incorrect answers decreased by 10%, partially correct answers barely varied (−5%), implying that some key concepts about climate change impacts were not fully understood. Despite this, the 15% increase in correct responses shows a positive overall trend in understanding this topic. However, it remains an area that requires further clarification in future workshops.

One of the most notable improvements was noted in Question 4, which concerned the historical and future pattern of land use change and its implications for water resources, as the incorrect responses dropped by 35% while the correct responses increased by 50%. This means that the dashboard was very useful for participants to appreciate that land use changes, particularly for land use directed at agriculture in a basin, change the amount and quality of water that is available in it. This result emphasizes the capability of the dashboard to translate complex data into comprehensible and applicable information for the vulnerability of water resources assessment.

Finally, for Question 5, which concerns the identification of the most vulnerable regions within the basin, there was a significant 45% decline in the incorrect answers and a 25% percent increase in the correct answers. Nevertheless, this has come up with some partially accurate responses, which indicate that some of the participants still do not understand the reason behind the regions’ differential vulnerability in the basin. This means that while the participants were more accurate in locating parts of the area where there was high vulnerability, the reasons as to why such areas were vulnerable are not clear enough, so it still needs to be addressed more comprehensively.

3.2. Usability and Display Effectiveness Evaluation

Figure 8 shows the distribution of ratings for each of the visualization effectiveness criteria evaluated in the vulnerability workshop.

According to the usefulness criteria, for most of the participants, the dashboard was a highly useful tool to understand the watershed’s vulnerability; 80% of the audience rated it 5 (according to the established Linkert scale). This indicates that the dashboard serves its role by offering relevant information and decision support. Regarding completeness, the information’s relevance was also highly rated, with 85% of participants giving the highest rating (5). This suggests that the information displayed on the dashboard is relevant and aligned with the needs and expectations of stakeholders in the context of vulnerability analysis. On the other hand, it also indicates that the dashboard’s content is well designed to address the main aspects of analysis in basin management with respect to water security.

Responses related to perceptibility and truthfulness showed more variability. While some participants found the information clear and reliable, a significant number did not (15% and 10%, respectively), suggesting that there is room for improvement in the way the information is presented, perhaps through more detailed explanations or the inclusion of glossaries and explanatory notes, as stakeholders mentioned in their feedback.

Even though the dashboard interface was judged as intuitive by most participants, 15% of them remarked that some parts could benefit from being more user friendly and confirmed that changes could be made to improve the user experience for those new to this type of tool. The esthetic experience was overall positive, with 70% of the participants scoring a 5. It shows how important the appearance for the perception of the tool is, and how much it can help to maintain user interest. The dashboard generated high interest from participants, with 80% scoring it with the highest rating (5). This is positive, as high interest can lead to a greater willingness to use the tool and apply the knowledge acquired for watershed vulnerability management.

Figure 9 compares the visualization evaluation profile developed for this research (green and purple profile) and the optimal profile proposed by Stephen Few (brown). The comparison reveals that, in general terms, the visualization is within the ranges considered adequate. However, it is observed that the veracity of the information is slightly below the optimal range. This finding suggests that it is necessary to implement an additional process of clarification with the basin stakeholders regarding the sources of information used. Although the data used come from official sources, time constraints in the participatory spaces have prevented a detailed analysis of each area of data, although a general description of their sources has been provided. It is important to note that to make this comparison, it was necessary to assign quantitative values to Stephen Few’s optimal profile scale, according to the scale established for this research. This is because this methodology is only qualitative.

On the other hand, it was considered pertinent to analyze the correlation between these evaluation criteria to determine the tool’s effectiveness. Figure 10 presents the results in a correlation matrix, revealing strong correlations between several criteria. For example, the perceptibility and truthfulness criteria correlated at 0.74, suggesting that participants who considered the information was clear also considered it was truthful, although the correlations with the other display efficiency criteria are moderate. Given the lack of correlation between the esthetics criterion and the other ones, it does not seem to be strongly correlated with the other criteria that were perceived by the participants (clarity, relevance, usefulness). On the other hand, the analysis shows that the criteria related to the usefulness, relevance, and interest of the dashboard have a high correlation, indicating that the stakeholders who found the dashboard useful also perceived it as relevant and able to generate interest.

The correlation matrix (see Figure 10) shows that there are no negative correlations because all the criteria analyzed are designed to evaluate complementary and positive aspects of the user’s experience with the dashboard. In essence, these criteria are neither competing nor mutually exclusive; on the contrary, they tend to be mutually reinforcing. For example, if a visualization is more perceptible (ease to understand visually), it is also likely to be perceived as more useful and truthful.

As for the presence of negative values in the scale, this is because from the theoretical sense of the concept of correlation, this could be negative. However, for the case under study, all these correlations are positive and indicate that, as one criterion improves (e.g., usefulness), it is more likely that others will also improve (e.g., truthfulness or interest). In summary, the matrix reflects the direct and constructive relationship between the criteria, suggesting that good performance in one criterion tends to be associated with good performance in the others.

3.3. Participation in Water Governance in a Transboundary Context

Based on the demographic characterization of the workshop participants, a significant bias in the composition of the attendees was identified, as most of them were from El Salvador, while the representation from Guatemala and Honduras was minimal. This may limit the construction of the regional perspective needed to effectively address the challenges associated with water resource management in the basin. Therefore, it is important to clarify that building relationships with water resource entities in the three countries in order to integrate them in the process of the co-creation of tools to address water security goes beyond the scope of the dashboard design objective.

However, this situation does not imply that the tool lacks contributions in the transboundary context of the basin. On the contrary, it facilitates the understanding of this territory as a unified system instead of three isolated and non-interconnected systems. In participatory spaces prior to the implementation of this research, a clear disaggregation of the basin according to the region corresponding to each country was identified. However, on this occasion, the CARL was conceptualized based on the sub-basins that compose it, even when several of these sub-basins extend across the three countries involved. This progress is significant for water governance and territorial management, as it promotes the development of strategies to address water security from a collaborative and articulated approach that benefits the system as a whole and avoids the duplication of efforts.

4. Discussion

Within the framework of the surveys conducted to evaluate interaction with the tool, there was an opportunity for the basin stakeholders to provide qualitative feedback. After a thorough analysis of the information collected, three main categories of information were identified: the strengths of the tool, challenges and limitations, and opportunities for improvement. The most relevant findings in each of these thematic areas are presented below.

4.1. Tool Strengths

• Interactivity and visualization: The possibility of direct and efficient comparison between different conditions and scenarios, thanks to the parallel visualization of maps and graphics, was highly valued. Therefore, this factor was considered as an enriching element by facilitating the interpretation of information through different perspectives. A better overall understanding was achieved by having visualizations that allow the information to be represented and analyzed in different ways. This is because the visualization that the dashboards facilitate leads to the expansion of more detailed information and is not limited to a single screen [51].
• Variables and structuring: The intuitive selection of variables was highlighted as a positive aspect. The user experience of the participants shows that the structure of the dashboard and the variables included facilitate interactive exploration with a rapid learning curve, conceptually assimilating what they represent in the model for the different scenarios.
• Specific areas of interest: The sections dedicated to demand, water quality, and vulnerability were identified as particularly useful by stakeholders. The usefulness of the water uses and the water quality indexes were also positively valued, as was the clear categorization of the different indexes. These elements provided a solid basis for analysis and allowed users to focus on critical areas for water resources management.

4.2. Difficulties in Its Use

• Interpretation of data and indexes: There was a partial perception of the indices as confusing and even counter-intuitive (especially the section related to failure frequency, probably because it is a concept with which there had been no previous interaction). In addition, difficulties were identified in understanding the index associated with groundwater behavior, suggesting the need for clarification. However, it is not entirely unexpected that there are difficulties in the interpretation of the information. One of the most frequent challenges in the use of dashboards is the misinterpretation of the information presented, which can result in not so accurate conclusions if the limitations in the data and their context are not adequately considered [51].
• Prior knowledge: Participants emphasized the importance of clearly understanding indicators, concepts, and weighting scales before using the dashboard. A lack of familiarity with these elements generated confusion during the analysis process, suggesting that providing educational resources or training sessions could significantly improve the user experience.
• Functionality limitations: A significant limitation noted was the inability to download data directly from the dashboard, which restricted the users’ ability to perform external analysis or custom presentations. Moreover, some stakeholders found the handling of percentages in the vulnerability analysis confusing, particularly regarding the failure frequency, which negatively affected the interpretation of the results.

4.3. Improvement Opportunities

• Training and support: Several stakeholders recommended increasing training on using the dashboard and its components. A section dedicated to practical exercises and examples of data handling and interpretation was suggested, with the objective of improving users’ ability to make informed decisions. This emphasizes the importance of providing additional resources to enable users to become thoroughly familiar with the tool and its applications.
• Additional functionalities: Among the proposals for improvement, the need for a more detailed glossary and the inclusion of the sources of the information used in the dashboard were highlighted. In addition, the ability to download data in formats such as spreadsheets was a functionality requested to facilitate external analysis. Another important suggestion was the incorporation of a comparative graph that would show historical and future scenarios simultaneously, avoiding the need to switch between scenarios on the map manually. Additionally, the generation of data at the micro-basin level and the inclusion of interpretative data scaled in the index tab were recommended to improve the understanding of results.
• Interface and user experience: The difficulty in the fluidity of the interaction experience with the dashboard was highlighted due to the speed of response of the dashboard. On the other hand, the inclusion of the legal context and criteria of each country was suggested to personalize and complement the analysis. Finally, it was recommended that explanatory information on the variables presented be included.

4.4. Contribution to Water Governance in a Transboundary Context

The results derived from the experience of using the vulnerability dashboard regarding the CARL, together with its detailed evaluation, represent a significant contribution to strengthening CARL water governance, considering its transboundary context. On the one hand, integrating historical information and projections of future uncertainties in an interactive and easily accessible environment through a robust tool facilitates an informed decision-making process, facilitating knowledge of the status and possible trends in terms of water security for the region [10]. This possibility enhances maneuverability in the face of growing challenges related to climate change and land use changes, particularly in agricultural contexts.

On the other hand, the tool makes it possible to identify vulnerabilities at the sub-basin level (not at the country level), allowing for the development of water resource management strategies at a more specific and detailed scale, as well as the identification of areas with a more critical situation; in this way, dashboards are a feasible tool for the characterization of insecurities in terms of water resources [52]. Additionally, it has proven to be a key element in strengthening the stakeholders’ understanding of the vulnerability of the CARL. By enabling stakeholders to experiment with thresholds and scenarios, the decision space visualization offers a practical method for evaluating adaptive strategies, ensuring their alignment with local water governance needs. Thus, such dashboards can be understood as a first approach to assessing water security for a subsequent more in-depth study of water resource vulnerability [10] for the identified areas.

Likewise, the evaluation of the interaction with the dashboard has revealed areas for improvement so that the tool can continue to be improved, and there is greater assurance in responding to the needs of stakeholders in the basin. This iterative feedback approach is critical to ensure that the dashboard remains a relevant and effective tool over time, because dashboards are more than communication tools—they are tools for exchanging feedback with stakeholders to improve engagement and the achievement of the benefits expected by them [51]. This provides the transfer of knowledge associated with this tool for long-term use, as well as its continuous updating.

It is also important to highlight how the dashboard reinforces a more participatory and inclusive governance model by facilitating a more active inclusion of the basin’s local stakeholders in the CARL vulnerability analysis process. The interaction with the dashboard promotes a sense of belonging to their region and co-responsibility in the management of water resources and reduces information asymmetry by providing easier access to information [51]. This is essential for effective regional water resources management and adaptation policies and strategies. Consequently, it is possible to ensure that in the face of present and future challenges to water resources, the response can be more resilient and adaptive.

In addition, the management of the CARL is strongly linked to its transboundary context, so it is essential to consider the inclusion of local stakeholders from the three countries that share this basin in the discussion related to its sustainability. The notable predominance of participants from El Salvador in the workshop is evidence of a bias that could limit the comprehensive understanding of the challenges facing the watershed; the homogeneity of the group evaluated in the implementation of the methodology [53] is often a recurrent limitation, which shows an option for improvement in future research. The imbalance in the nation’s participation suggests a need for intentional engagement with other countries for more equitable participation and input in informing watershed-level transboundary water security strategies.

However, the use of the vulnerability dashboard has allowed for a more inclusive and unified conceptualization of the CARL, facilitating the visualization of these from the sub-basins that compose it and their interconnection. In this way, in addition to promoting articulation and collaboration between countries, the establishment of more constructive dialogs in the approach to water security is promoted, fostering greater commitment on the part of local stakeholders in the regional management of water security [10], resulting in the development of more effective strategies. For these reasons, to ensure that all the voices of local stakeholders are considered in the co-creation of management tools to respond to the needs of the entire basin, it is necessary to continue strengthening the articulation between the entities of Guatemala, Honduras, and El Salvador. It is evident that in today’s data-driven world, the integration of complex information into interactive graphs and metrics enables stakeholders to quickly and easily access critical data to identify patterns and trends across scenarios for water management, driving improved decision-making [9]. Thus, while making it easier to interact with data, it also promotes informed decision-making based on reliable information geared towards sustainability and equity in water resources management.

5. Conclusions

The obtained results confirm that the dashboard is an effective tool for visualizing and analyzing the vulnerability of water resources of the CARL to different future uncertainties, such as climate change and agricultural land use change, providing a robust framework for informed decision-making. The evaluation of the dashboard, using a mixed approach that combined pre- and post-use knowledge testing with an assessment of stakeholder experience, validated its usefulness in both conveying key concepts and strengthening users’ capacity to manage water resources more effectively.

On the one hand, the results of the pre- and post-use knowledge tests show that the use of the dashboard implied a notable improvement in the participants’ knowledge of the main factors affecting the basin’s water vulnerability, being most effective in areas related to water uses and land use change. This shows the potential of dashboards, not only for the capacity building of local stakeholders, but also to support their engagement in participation in the development of adaptative governance strategies, especially through the identification of gaps in knowledge about water resource vulnerability that must be addressed in participatory workshops. However, issues such as the multiple factors influencing vulnerability and the most vulnerable regions still require further clarification. This suggests that there are areas of improvement for the development of the dashboard, which could be addressed in future interactions in workshops such as the one already implemented or with the adjustment of the dashboard interfaces and content.

Additionally, the evaluation of the interaction with local stakeholders reflected the usefulness and relevance of the dashboard’s interactive visualization and adaptative functionalities, as well as its visual appeal, facilitating the interpretation of complex data and the ability to compare scenarios, both spatially and temporally. However, usability challenges were identified, particularly for less technically experienced users, suggesting the need for a training-focused approach; for example, we propose the following: incorporating tutorials for use, guides for the interpretation of concepts and indicators, detailed glossaries, and pop-up windows with contextual information on variables and indicators, among others. These challenges underscore the importance of the ongoing optimization of the dashboard’s interface and functionality to make it accessible to a wider range of users. It is essential to address these usability challenges as an enhanced accessibility and understanding of the tools will directly impact the willingness and capacity of users (policy-makers and local stakeholders) to adopt informed and effective water governance policies.

At the water governance level, the dashboard has become an important means for approaching local stakeholders in the decision-making process, creating a demand to understand the threats and vulnerabilities of the basin. This tool allows for an effective sharing and participatory approach to water management which is important for a transboundary region like the CARL, which covers El Salvador, Guatemala, and Honduras. Water sustainability in the basin requires a process of coordination and cooperation among all countries, and the dashboard facilitates this process by bringing data to light and enhancing its accessibility to all stakeholders. This not only enhances transparency, but it also promotes the consolidation and ownership of the region by enabling the identification of sub-basin vulnerabilities as well as decisions to be made based on common evidence.

There is need to establish more balanced collaboration mechanisms to facilitate a more equitable representation in the transboundary context of the basin, in accordance with the observed bias of participation relative to the composition of the workshop. This is critical for the approaches and solutions formulated for water security to be inclusive and sustainable, in harmony with the realities and requirements of the entire basin. This will lead to sound water governance that comprehensively considers and addresses the interrelated challenges facing the CARL and springs, creating a more resilient and integrated future for managing its water. But until now, the tool has encouraged the visualization of the basin as a unified system, not one divided by national borders. This outlook encourages the development of water security strategies in a concomitant and non-overlapping manner. It therefore aims to resource use optimization more efficiently, considering the interest of all countries involved in managing the basin.

For future research, it is essential to extend the vulnerability analysis of the CARL by incorporating additional scenarios, such as socioeconomic changes and the effects of water management policies and strategies, to promote water security in this transboundary region. Also, adopting an equity and accessibility approach is key to evaluating how different social and economic groups within the basin access visualization tools, as well as identifying the barriers that hinder their access to critical information. On the other hand, it would be valuable to develop a study on the long-term impact of the tool in its capacity to influence the creation or modification of water governance policies and practices through the development of quantitative and qualitative indicators that are adaptable to different contexts and scales of analysis.

The key issues to address to ensure that the dashboard fosters greater transparency and public participation in governance and is used by decision-makers to improve the management of water resources in this watershed are the following: (1) the incorporation of user feedback and the inclusion of adaptation strategies and their impacts in the dashboard; (2) the commitment and training of a local institution to host and maintain the dashboard and to make it available the public; and (3) the engagement of decision-makers of the three countries in the use of the dashboard to understand future uncertainties and the potential impact of adaptation strategies on performance metrics.

In conclusion, this research has important contributions for water sources management and governance, particularly in a transboundary context such as the CARL. Through this study, the potential of dashboards as interactive support tools in decision-making processes to improve transparency, accessibility, and collaboration between local stakeholders from different regions and levels of governance is evidenced. This interactive nature of the dashboard promotes the participation of local stakeholders to increase their sense of ownership and shared responsibility in water governance. Furthermore, through its evaluation, it was possible to identify how the dashboard facilitates the identification of the most vulnerable sub-basins to prioritize their management and promotes a shared understanding of the challenges of adaptation strategies to these vulnerabilities (based on evidence and scenario analysis). All these findings highlight the need for continued capacity building, institutional commitment, and regional cooperation with local stakeholders. In this way, it will be easier to ensure better long-term management and the integration of this type of tool into approaches to water management and policy development concerning water resources.