Assessment and spatial mapping of water quality parameters using QGIS: Creating a dynamic map Open Access

Groundwater, a crucial natural resource, serves as a viable alternative to surface water for drinking, irrigation, and industrial purposes (Arslan 2012; Jang et al. 2013; Thlakma et al. 2018). However, groundwater contamination leads to several critical problems, including poor drinking water quality, high water purification costs, human health concerns, and depletion of water supply. To address these issues, it is essential to safeguard groundwater quality through effective planning strategies. This includes meticulously monitoring the chemical, physical, and biological conditions of the groundwater (Jang et al. 2013). Notably, groundwater, constituting 30.1% of global freshwater resources, emerges as the second-highest source after glaciers and ice caps, accounting for 68.7% (Water Science School 2018; Shiklomanov (1993).

In Malaysia, the overall production of treated groundwater in Peninsular Malaysia and the Federal Territory of Labuan reached 201 million L/day (1.2%) in 2020 (Malaysia Voluntary National Review (VNR) 2021). This highlights the potential for enhancing the utilization of groundwater to achieve a more balanced distribution among various water sources.

Groundwater quality assessment plays a crucial role in addressing potential human health challenges arising from exposure to toxic contaminants across diverse environmental media. Its objective is to estimate the severity or magnitude of the risk to human health posed by exposure to environmental hazards (Achour et al. 2005; Wen et al. 2006; Thlakma et al. 2018). In recent years, there has been a surge in research focused on spatial and spatiotemporal modelling of water and groundwater quality. This increasing interest is fuelled, in part, by advancements in computational power, remote sensing, earth-bound surveys, and cartography. These developments have led to the emergence of integrated mapping tools, commonly known as geographic information systems (GIS) (Burrough et al. 1998; Benalcazar et al. 2024), which will be utilized in this research to draw maps and use a geostatistical technique to interpolate the groundwater quality index (GWQI) in Peninsular Malaysia. Geostatistics refers to a branch of spatial statistics that characterizes spatial patterns and provides estimates of attribute values at unsampled locations. It offers a comprehensive framework for integrating diverse datasets to create localized models of spatial uncertainty. Numerous studies showcase the application of various geostatistical techniques to address groundwater pollution issues (Panagiotou et al. 2022). Other studies have employed quantum GIS (QGIS), a freely accessible GIS, as a powerful tool for the creation and analysis of spatial and spatiotemporal models in the domain of groundwater quality. The versatility and user-friendly features of QGIS, coupled with its cost-free availability, have made it a preferred choice among researchers (Batarseh et al. 2021; Kpiebaya et al. 2022; Bennett 2023; Ogarekpe et al. 2023).

The lack of recent studies in Peninsular Malaysia addressing the dynamic mapping of groundwater quality poses a significant challenge. Traditional approaches involve deploying an extensive network of groundwater stations or wells to establish groundwater quality data. However, the conventional methodology is time-consuming and resource-intensive. Considering this, adopting QGIS interpolation for creating a spatial map of the GWQI emerges as a more advantageous and efficient alternative. By employing spatial interpolation techniques, researchers and policymakers gain the capability to discern groundwater quality at specific points within Peninsular Malaysia as preliminary data. The study aims to generate a spatial distribution for the GWQI, perform a spatiotemporal assessment of the GWQI across the study area from 2014 to 2022, and develop a dynamic map that ensures easy public access, facilitating a user-friendly interface for individuals to interact with and explore the GWQI map.

Previous studies used similar indexes such as the heavy metal pollution index (HPI) for assessing heavy metal contamination in groundwater (Zainudin et al. 2023), entropy water quality index (EWQI) for evaluating groundwater suitability (Li et al. 2021; Saraswat et al. 2023), GIS-based GWQI for spatial analysis (Machiwal et al. 2011), and the drinking water quality index (DWQI) (Ramesh et al. 2010), and artificial neural network for predicting Escherichia coli presence (Khan et al. 2021), as summarized in Table 1.

Many studies related to water and groundwater quality have used the GIS for mapping and interpolation. A study focused on mapping groundwater potential using QGIS and a multi-criteria decision analysis technique of Analytic Hierarchy Process (AHP), integrating hydro-geophysical surveys and recharge estimations for deeper insights (Kpiebaya et al. 2022). Another study in Iran utilized geostatistical techniques and GIS interpolation to prepare agricultural quality plots based on hydro-geochemistry (Sheikhy Narany et al. 2014; Mageshkumar et al. 2023). In India, GIS was employed at a 1:50,000 scale in the Nagapattinam district to create foundational maps and analyse hydrological parameters and water quality using ArcGIS spatial analyst tools (Gnanachandrasamy et al. 2015; Mageshkumar et al. 2023). Similarly, GIS was used to evaluate the water quality of River Chittar and connected systems in South Tamil Nadu, generating maps over a 2-year period to assess various parameters (Drusilla et al. 2004; Mageshkumar et al. 2023). Other case studies were carried out in India for the investigation of groundwater (GW) for domestic and irrigation quality mapping (Remesan & Panda 2007; Satyanarayanan et al. 2007; Gupta & Srivastava 2010; Nas & Berktay 2010).

Dynamic maps are highlighted as crucial tools in geospatial technology, offering an interactive means to visualize and analyse spatial information (Roth 2013). These maps enhance spatial analysis and decision-making processes, facilitate collaboration and knowledge sharing, and are integral to various domains due to increased accessibility and mobile mapping capabilities (Hossain & Meyer 2018). Additionally, studies have utilized dynamic maps in groundwater quality assessment, such as leveraging the DRASTIC index (DRASTIC index is an index that considers seven parameters namely; Depth to water, Recharge, Aquifer media, Soil media, Topography, Impact of vadose zone media, and Hydraulic Conductivity) in QGIS to depict groundwater vulnerability spatially (Duarte et al. 2015), and developing interactive map services by governmental agencies to provide comprehensive water resources data essential for informed decision-making (Miami-Dade County Government 2021; GEMS/Water Data Centre 2014; Government of Newfoundland and Labrador 2023).

Geostatistical interpolation methods like ordinary kriging (OK) and inverse distance weighted (IDW) assume that closer values share greater similarity. While IDW predicts values based solely on distance, it is considered accurate for spatial interpolation by using known values and corresponding weighted values.

A study in Richards Bay, South Africa, employed spatial interpolation and statistical methods using OK and IDW, revealing that OK better maps highly contaminated areas surrounded by less contaminated ones, precisely identifying three contaminated zones by landfill, industries, and agriculture, while IDW connects all contaminated zones and labels most of the study area as contaminated (Elumalai et al. 2017). In another study, OK was used to estimate and map the spatial distribution of electrical conductivity (EC) in a county, revealing potential salinization areas in the northern and southern parts, with recommendations for soil salinization prevention measures, and observing nitrate pollution attributed to runoff and excessive fertilizer use in county-seat areas (Hu et al. 2005). Conversely, a study in Texas explored interpolation methods for estimating arsenic levels in groundwater, finding IDW exhibiting a higher correlation coefficient with measured arsenic levels compared to Kriging, especially in different regions and aquifers, with regression analysis emphasizing the importance of including well depth and elevation as covariates for accurate estimation, particularly with IDW (Gong et al. 2014). Additionally, IDW interpolation was utilized in the Khambhat coastal region of Gujarat, India, to create spatial maps depicting the geographical distribution of water quality parameters and groundwater quality status based on the GWQI (Bhavsar & Patel 2023). Moreover, a hypothetical case study applied OK to predict total organic carbon levels in observations from the Passaic River in New Jersey, highlighting the limitations of OK in estimating spatial distribution due to uniform sampling assumptions (Zhou & Michalak 2009).

The study utilized the Malaysia GWQI, which is constructed based on seven parameters: pH, iron, total dissolved solids (TDS), nitrate, sulphate, phenol, and E. coli. It employs a quality scale ranging from 0 to 100. This scale helps identify the groundwater quality, ranging from very polluted to very good. The GWQI is built based on an arithmetic approach; weighting factors are introduced for each parameter, as shown in Equation (1).

If the numerator (concentration of the sub-index in mg/L except for E. coli most probable number (MPN)/100 mL) exceeds its denominator, is set to 0, where Si stands for sub-index.

Data for groundwater quality assessment were obtained from the DOE's Environmental Quality Reports (Department of Environment 2018, 2022), encompassing a comprehensive dataset calculated to derive the GWQI. The dataset comprises information from 89 groundwater quality monitoring stations (wells) located across Peninsular Malaysia, offering a representative coverage of the region. The temporal scope of the data spans from 2014 to 2022, providing a multi-year perspective for the assessment of groundwater quality dynamics.

The study used WGS 84 as the QGIS project coordinate reference system (CRS). A Google Terrain Hybrid basemap was used as the base layer through the HCMGIS plugin, and a polygon layer outlining the boundaries of Peninsular Malaysia was obtained from the Humanitarian Data Exchange website (OCHA Regional Office for Asia and the Pacific (ROAP) (2023)) to define the study area boundaries. Collecting the GWQI data was first compiled into a CSV file and then transformed into a multipoint vector layer. IDW interpolation was selected over OK due to limitations in the QGIS software. Specifically, OK requires a different CRS, particularly Universal Transverse Mercator (UTM), to function in larger study areas. However, since UTM has multiple zones, each with its own CRS, this presented a challenge: the study area spans two UTM zones, N47 and N48, which the Kriging plugin does not support for multiple-zone interpolation. Consequently, IDW was chosen as a suitable alternative, with previous studies indicating that IDW often achieves a higher correlation compared to Kriging in larger areas, as discussed in the literature. IDW was employed to generate a raster layer depicting the spatial distribution of the GWQI values across the study area. This process was repeated for each year within the study period to capture the GWQI variations over time. Finally, raster layer unique value reports were extracted to analyse the GWQI distribution patterns, and maps were exported to visually represent the spatial distribution of the GWQI for each year.

The QGIS2Web plugin was used to craft an interactive and dynamic map. This research harnessed the QGIS2Web plugin's capability to export QGIS projects to OpenLayers or Leaflet webmaps, reproducing project elements such as layers, extent, and styles, all without necessitating server-side software. It is noteworthy that the plugin is freely available, adding to its accessibility and user-friendly attributes.

Following the completion of spatial mapping in QGIS, the QGIS2Web plugin seamlessly transformed the project into a webmap. In this specific study, the decision was made to export the project to a Leaflet webmap due to its simplicity and lightweight characteristics. Subsequently, the resulting webmap was uploaded to a hosting server, improving accessibility by allowing users to interact with the dynamic GWQI maps without the need to download the map. This approach ensured a smooth and user-friendly experience for accessing and exploring the dynamic GWQI maps online.

The majority of Peninsular Malaysia consistently displays good groundwater quality throughout the period under study. Green areas dominate the maps, suggesting stable and generally favorable groundwater conditions across most regions. In contrast, medium-quality areas are more variable, with the presence of yellow regions fluctuating over the years. This variability points to localized issues impacting groundwater quality, necessitating targeted intervention and management. Areas with very good groundwater quality are sparse, with few regions consistently showing blue. This indicates that optimal groundwater conditions are limited and concentrated in specific regions.

From 2014 to 2016, the initial maps show a strong presence of green (good quality) areas with some yellow (medium quality) regions in the northern and central parts. By 2016, there is a noticeable reduction in yellow areas, indicating an improvement in groundwater quality in these regions. During the period from 2017 to 2019, the central region sees a resurgence of yellow areas, signifying a decline in groundwater quality. The southern regions also show sporadic medium-quality areas, reflecting localized issues impacting groundwater. In the years from 2020 to 2022, the later maps indicate a mixed trend. While some central areas show improvement (reduction in yellow areas), other parts of the central and southern regions exhibit increased medium-quality areas. The northern regions largely maintain good quality, with occasional fluctuations but generally showing stability.

Regionally, the northern region demonstrates relatively stable and good groundwater quality over the years, with occasional medium-quality areas that generally show improvement. The central region is characterized by significant variability, with alternating periods of decline and recovery in groundwater quality. The southern region shows more frequent changes, with periods of medium-quality areas appearing and then improving.

In 2014, the groundwater quality in Peninsular Malaysia was predominantly good, with 94.92% of the area falling within the 70–90 GWQI range. Medium-quality areas (40–70 GWQI) accounted for 3.09%, while only 1.99% of the area had very good-quality groundwater (90–100 GWQI). The subsequent years saw some variations. In 2015, the area with good-quality groundwater increased to 99.30%, with a corresponding decrease in both medium and very good-quality areas to 0.36 and 0.34%, respectively. This trend of high proportions of good-quality groundwater continued, though with minor fluctuations, until 2017.

However, 2018 marked a significant change with an increase in medium-quality areas, which reached 12.67% and a decrease in good-quality areas to 87.04%. This indicates a substantial decline in groundwater quality, which persisted into 2019 with 10.30% of the area classified as medium quality and 89.62% as good quality. The presence of very good-quality areas remained minimal throughout these years, never exceeding 1.99%.

From 2020 to 2022, the groundwater quality showed some recovery. In 2020, the good-quality area rose to 94.76%, with medium-quality areas decreasing to 4.40%. This trend continued into 2021 and 2022, with good-quality areas at 94.47 and 92.0%, respectively. Despite these improvements, the very good-quality classification remained almost negligible, indicating that optimal groundwater conditions are still limited and localized.

Over the 8-year period, the area covered by the GWQI 40–70 (medium quality) classification showed a significant increase of 155.66%. Conversely, the area covered by the GWQI 70–90 (good quality) classification showed a decrease of 3.08%. The most notable change was observed in the area covered by the GWQI 90–100 (very good quality) classification, which decreased by 98.49%.

Analyzing the average, minimum, and maximum values in Table 4 over the entire period provides additional insights. The average area percentage for medium quality (40–70 GWQI) was 5.09%, with a minimum of 0.36% in 2015 and a maximum of 12.67% in 2018. For good quality (70–90 GWQI), the average was 94.37%, with a minimum of 87.04% in 2018 and a maximum of 99.30% in 2015. The very good-quality classification (90–100 GWQI) had an average of 0.54%, with a minimum of 0.03% in 2022 and a maximum of 1.99% in 2014.

A dynamic web map of the interpolated GWQI data for Peninsular Malaysia has been created using the QGIS2WEB plugin and is hosted on a GitHub repository (https://ebrahimaaj.github.io/Groundwater-Quality-Index/). This interactive map offers several functionalities for users to explore the spatiotemporal patterns of groundwater quality.

The dynamic map is optimized for a smooth user experience with a file size under 10 MB, achieved through station clustering and reduced raster resolution. This optimization facilitates faster loading times and efficient data exploration.

The map functionality empowers users to conduct a variety of analyses. Users can identify areas with potentially problematic groundwater quality (hotspots), track changes in the GWQI over time, and investigate potential correlations between the GWQI and other environmental factors. For instance, users can compare groundwater quality near solid waste landfills with golf courses. This dynamic web map offers significant practical value for stakeholders, researchers, and policymakers involved in groundwater quality management. The interactive exploration of spatial and temporal trends in the GWQI can enhance understanding of how surrounding environmental factors influence groundwater quality. This knowledge can inform decision-making processes related to groundwater resource management, pollution mitigation strategies, and the development of sustainable practices for protecting groundwater resources in Peninsular Malaysia.

In conclusion, this study generated the spatial distribution of the GWQI for 89 stations across Peninsular Malaysia covering the period 2014–2022, using QGIS software and the IDW interpolation method. This research suggested that there are no GWQI values lower than the medium classification (40–70), indicating that the lowest GWQI in Peninsular Malaysia is suitable for raw water, industrial use, irrigation, and agricultural use. It also revealed that the majority of Peninsular Malaysia consistently displays good GWQI classification (70–90) throughout the period under study. Additionally, areas with very good GWQI classification (90–100) are sparse, indicating that optimal groundwater conditions are limited and concentrated in specific regions.

This study performed a spatiotemporal assessment of the GWQI across the study area from 2014 to 2022 using QGIS and its raster layer unique value report. The assessment identified deteriorations in groundwater quality over time, with areas of the medium GWQI classification increasing by 155.66%, while areas of very good GWQI classification decreased by 98.49%. These findings highlight the need for continuous monitoring and targeted interventions to address groundwater quality deterioration. The 2014–2022 period provides a recent view of groundwater quality, highlighting short- to medium-term changes due to land use or industrial activities. However, this timeframe may limit the detection of long-term trends, as gradual processes and cumulative impacts often emerge over decades. While useful for understanding current conditions, a longer dataset would offer deeper insights into sustained trends.

Furthermore, this research developed a dynamic, user-friendly map for public access using QGIS and the QGIS2WEB plugin. This map assists individuals, researchers, and policymakers in exploring groundwater quality, identifying areas with potential issues, tracking changes in the GWQI over time, and investigating correlations with other environmental factors. The interactive web map offers a practical value for stakeholders in groundwater quality management, enhancing their understanding of environmental influences on groundwater quality and supporting decision-making for resource management, pollution mitigation, and sustainable practices. This initiative aligns with Malaysia's 21st-century water vision, which seeks to raise public and leadership awareness of internal and external trends affecting future water use, thereby strengthening the political commitment and insights needed to guide a strategic action framework (Food & Agriculture Organization (FAO) 2024). Additionally, similar to other studies, this study aims (Makubura et al. 2022) to provide information for policy- and decision-makers as well as general awareness. It also helps with surface and groundwater monitoring.

Further investigations are needed to comprehensively improve groundwater quality monitoring and management in Peninsular Malaysia. First, increasing the number of the GWQI monitoring stations and ensuring their uniform distribution across the region is essential for better accuracy and reliability of the spatial representations of groundwater quality. Additionally, it is crucial to investigate the external factors contributing to groundwater quality deterioration. Understanding these factors, such as industrial activities, agricultural practices, and urbanization, will provide insights into the sources of contamination and help in developing targeted mitigation strategies. Moreover, examining the each GWQI parameter in detail and generating a spatial distribution is necessary to understand the specific changes in groundwater quality over time. This detailed analysis will help identify the underlying causes of quality fluctuations and guide the implementation of appropriate interventions to safeguard groundwater resources.

By addressing these limitations and following the proposed recommendations, future research can substantially improve the understanding of groundwater quality in Peninsular Malaysia. This knowledge can inform effective strategies for groundwater resource management, pollution mitigation, and the development of sustainable practices to protect this vital resource. Additionally, the scalability of this approach to other regions or countries further enhances its impact. With the adaptability of open-source tools like QGIS, similar methodologies can be applied across various geographies, accommodating local environmental conditions, regulatory standards, and data availability. This potential for broader application underscores the role of accessible technology in advancing sustainable groundwater management on a global scale.

The authors would like to acknowledge Universiti Tenaga Nasional (UNITEN) and the University of Nizwa.

E. A. and G. H. conceptualized the study, wrote, reviewed, and edited the article.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.