Linking Urban Sustainability and Water Quality: Spatial Analysis of Topographic, Sociodemographic, and Flood-Related Factors Affecting Well Water in Jakarta (2017–2019)

Abstract

In 2019, well water was the primary source of clean water for 76.18% of Indonesian households. In the same year, based on an assessment of water quality in subdistricts in Indonesia, Jakarta had the third-lowest water quality index. This research aimed to analyze the impact of topographic, sociodemographic, and flood-related factors on well water quality in Jakarta from 2017 to 2019. This study employed an ecological design and used data obtained from various government agencies and has been published on an official website. Water quality data from wells in 261 subdistricts were analyzed using correlation and spatial analyses in this study. More than 83% of water well quality in Jakarta failed to meet the standard within the period of the study. The well water quality was the poorest in North Jakarta with the lowest elevation above sea level. The factors significantly associated with well water quality were low elevation (p ≤ 0.001), high population density (p = 0.015), and a low education level (p = 0.028). Local governments and private sectors should expand the piped water network and educate the public on well water quality and how to prevent waterborne diseases.

1. Introduction

Water quality has a direct effect on the sustainability of groundwater resources for human use [1]. Water must meet quality standards, as poor water quality can have adverse health effects (e.g., diarrhea and dysentery hepatitis A, typhoid fever, legionellosis, polio, trachoma, and schistosomiasis) [2]. Effective management strategies, which include the long-term monitoring of the effects of natural and anthropogenic factors on water quality, are vital for ensuring a sustainable water supply in regions such as the High Plains aquifer in the United States [3].

As of 2019, the majority of households in Indonesia (76.18%) relied on groundwater as their primary source of clean water, with nearly half (48.19%) using groundwater as drinking water [4]. Currently, Indonesia faces a number of issues related to water quality due to an insufficient piped water supply [5], resulting in increased groundwater consumption in the country, particularly in the capital, Jakarta [6].

Despite a rising demand for water in Indonesia, the availability and quality of groundwater have declined. In 2019, Jakarta had the third-lowest water quality index after Yogyakarta and East Nusa Tenggara [7]. In 2018, monitoring by the Environmental Agency of 267 groundwater sampling points across Jakarta revealed that 40.4% of the groundwater was slightly polluted, 34.5% moderately polluted, and 11.2% heavily polluted [8].

The risk factors associated with changes in well water quality include land use changes [9], population density [10], improper industrial waste management [11], education levels sanitation facility conditions [12], government policies [13], land subsidence [14], rainfall patterns [15], and flooding [16]. Among these factors, land use is one of the major causes of groundwater quality changes [9], often influenced by population growth and economic development. Urbanization and industrialization increase pollution through pesticides, fertilizers, and household waste [17,18]. A study in India found that higher population density significantly impacted water quality, particularly increasing NO₃ and K concentrations [19].

Rapid population growth that is not matched by land availability results in limited sanitation facilities in residential areas, leading to water quality deterioration due to pollution from household activities [20]. Sanitation facilities contribute to changes in groundwater quality, including well water. A study in Africa showed that as the density of sanitation facilities, particularly latrines (ground toilet facilities with limited storage/no treatment) increased in an area, nitrate and chloride levels in unprotected wells significantly increased [12]. Jakarta, with a population of approximately 11 million, continues to face pollution challenges stemming from inadequate wastewater treatment. Thus, approximately 1.3 million residents (11%) still discharge domestic wastewater directly into rivers without treatment [21]. Moreover, as of 2021, access to minimally contaminated piped water remained low (8%) [22].

Rapid industrialization often occurs without careful planning and compliance with regulations, particularly in terms of waste management. Groundwater pollution can occur via various routes, such as leaking storage tanks and waste disposal pipes due to poor maintenance or construction, as well as accidents like fires and explosions [23]. Common industrial pollutants causing groundwater contamination include heavy metals, organic solvents, hydrocarbons, chromium, nitrogen, sodium chloride, and chlorinated solvents [23].

Education level influences well water quality, as individuals with higher education are more knowledgeable about pollution prevention and sanitation standards [5,24]. They are also more likely to invest in sanitation facilities [25] and have better access to piped water at the community level [5]. Higher education enhances health awareness, leading to improved sanitation practices and a greater willingness to invest in health and water quality [5,26]. A previous study highlighted that educated individuals are more willing to pay for better health and sanitation [5].

Geological conditions can also affect water quality. Land subsidence due to changes in land use and excessive groundwater exploitation can affect water quality. Subsiding land impacts groundwater quality through seawater intrusion and groundwater pollution [14]. Urbanization and land use changes alter topography, creating phenomena that disrupt recharge processes and affect groundwater quantity and quality [27].

Intense rainfall over short periods can enhance water flow and facilitate the transport of pathogens from contamination sources, such as latrines and waste dumps, into groundwater via aquifer pathways [15].

Flooding causes pollutants from various sources to flow into waterways, resulting in contamination [16]. The capacity of water drainage systems evolves over time, often influenced by human activities, such as unprocessed domestic waste, industrial wastewater disposal, and land use changes that reduce river widths [28]. As a result of this reduced capacity, during precipitation events, water carrying pollutants flows into water channels and rivers [16].

Jakarta experiences annual flooding, with flood heights reaching 3–9 m in 1993–2007, affecting 56–159 cities [29]. Research pointed to a decline in river water quality in 2008–2014 in the province based on physical parameters, such as biochemical oxygen demand and total suspended solids, at most monitoring points. A study in Thailand found that flooding in 2011 contaminated public tap water with disease-causing bacteria, such as Shigella, Vibrio cholerae, and Leptospira, due to runoff from agricultural areas and fecal contamination [30].

Flooding can contaminate water sources with waste and pathogens, especially in areas with limited sanitation infrastructure [31]. The utilization of spatial technology enables effective risk mapping and targeted interventions, ensuring the resilience of clean water, particularly for communities with vulnerable socio-economic conditions [32].

In context, this study aims to analyze the relationship between topographic factors (elevation and rainfall), sociodemographic factors (population density and education level), and flood events and well water quality in Jakarta from 2017 to 2019 with a view to understanding how these factors contribute to the challenges of maintaining water quality for sustainability in urban areas.

2. Materials and Methods

2.1. Data Resources

This study employed an ecological design and a spatial analysis approach. The dependent variable was well water quality. The independent variables consisted of topographic factors (elevation and rainfall), sociodemographic factors (population density and education level), and environmental factors (flooding). The unit of analysis was individual subdistricts (N = 261) in Jakarta. This study was approved by the Faculty of Public Health Universitas Indonesia Ethical Board on 21 May 2024 (approval number: Ket-350/UN2.F10.D11/PPM.00.02/2024). All the data used in this study were obtained from secondary sources from various agencies. The topographic data, including elevation information, were obtained from the Indonesia Geospatial Portal. The rainfall data were sourced from NASA (The NASA POWER Project’s Data Access Viewer) and Meteorology, Climatology, and the Geophysics Agency (Badan Meteorologi, Klimatologi, dan Geofisika, BMKG). The flood water level data were acquired from the Regional Disaster Management Agency of Jakarta. This study was conducted from April 2024 to June 2024.

2.2. Data Analysis Software

In this study, data analysis was performed using several software tools. SPSS version 25 was used to analyze frequency distribution, mean, mode, median, and other descriptive statistics for each variable. Additionally, Stata version 17 was applied to conduct Spearman correlation tests between independent and dependent variables. For spatial data visualization, ArcGIS version 10.8 was used to create choropleth maps, graduated symbol maps, and perform overlay analysis.

2.3. Proportion of Residents in Each Subdistrict with a Low/High Educational Level

Educational attainment data were obtained from the Satu Data Jakarta Portal, uploaded by the Population and Civil Registration Office of Jakarta. Education was classified into two levels: low and high.

• Low education level: Residents whose education was incomplete (i.e., completed only elementary school or junior high school);
• High education level: Residents whose education was complete (i.e., completed high school or university).

A basic formula for demographic and statistical analysis can be used to calculate a proportion that gives the percentage of the residents in each subdistrict with a low or high education level [33]. The formula used to calculate the proportion of residents in each subdistrict with a low or high education level was as follows [33]:

$${\displaystyle \frac{\sum \mathrm{resident} \mathrm{with} \mathrm{low} \mathrm{or} \mathrm{high} \mathrm{education} \mathrm{level}}{\sum \mathrm{resident} \mathrm{per} \mathrm{subdistrict}}}\times 100\%$$

(1)

2.4. Population Density

Population density data were obtained from the Satu Data Jakarta Portal, uploaded by the Population and Civil Registration Office of Jakarta. Population density was defined as the number of residents per unit area. The formula used to calculate population density was as follows [34]:

$${\displaystyle \frac{\sum \mathrm{residents} \mathrm{per} \mathrm{subdistrict}}{\mathrm{Area} \mathrm{of} \mathrm{the} \mathrm{subdistrict}}}$$

(2)

The resulting population density figures were classified into four categories (

Table 1. Classification of population density.

Classification	Population Density (People/km2)
Low	<15,000
Medium	15,000–20,000
High	20,001–40,000
Very high	>40,000

Source: Badan Pusat Statistik, 2009 [34].

).

2.5. Water Pollution Index (WPI)

To calculate the WPI, water quality parameters were determined, according to Regulation Number 27 Year 2021 of the Ministry of Environment and Forestry of the Republic of Indonesia [35]. In total, nine water quality parameters, namely total dissolved solids (TDS), turbidity, iron (Fe), manganese (Mn), nitrate, nitrite, pH, total coliform, and Escherichia coli, were assessed to determine the groundwater pollution load and WPI based on permissible standard limits defined by the Ministry of Health Regulation Number 2 of 2023 concerning the implementation of government Regulation Number 66, 2014 concerning Environmental Health [36]. The WPI was calculated using the following formula [35]:

$$\mathrm{IPj}=\sqrt{{\displaystyle \frac{{\left\{{\left(C\right\}}_i/ L_{ij}\right)}_M^2+{\left\{{\left(C\right\}}_i/ L_{ij}\right)}_R^2}{2}}}$$

(3)

In the formula

• IP_j—pollution index value;
• C_i—concentration of the i-th pollutant in the water sample;
• L_ij—permissible concentration limit for the i-th pollutant as per relevant water quality standard;
• M—maximum concentration ratio C_i/L_ij observed among all the parameters;
• R—average concentration ratio C_i/L_ij observed among all the parameters.

The obtained WPI pollution values were then interpreted according to the established standards for pollution levels (

Table 2. Interpretation of Water Pollution Index Values.

Water Quality Interpretation	Water Pollution Index (WPI) Value
Good (Meets Standards)	0 ≤ IPj * ≤ 1.0
Slightly Polluted	1.0 < IPj ≤ 5.0
Moderately Polluted	5.0 < IPj ≤ 10.0
Heavily Polluted	IPj ≥ 10.0

* IPj, pollution index value. Source: Regulation Number 27 Year 2021 of the Ministry of Environment and Forestry of the Republic of Indonesia [35].

).

2.6. Data Analysis

Statistics analysis was performed using correlation tests because both the dependent and independent variables were ratio-scaled. The strength of the relationship between the two variables was assessed using the correlation coefficient (r value). Spatial analysis was conducted using a range of methods (e.g., choropleth mapping, graduated symbol mapping, and overlay analysis) to visualize the distribution of the independent and dependent variables in each subdistrict in Jakarta.

3. Results

3.1. Distribution of Water Pollution in Jakarta from 2017 to 2019

Drawing on data compiled by the Jakarta Environmental Agency [37], for each water quality parameter, water quality was determined in two periods: the dry season and the rainy season. Based on the analysis of the water quality parameters, the average concentration of each parameter was higher during the rainy season than during the dry season in all years, except for the total coliform concentration.

Table 3. Well water quality parameter and average concentration in Jakarta from 2017 to 2019.

Water Quality Parameters	Average Concentration		Mean	Quality Standards 1
	Dry Season	Rainy Season
Physical
TDS (mg/L)	448.4	473.48	460.94 *	300
Turbidity (NTU)	3.35	3.44	3.4 *	3
Chemical
Fe (mg/L)	0.13	0.16	0.15	0.2
Mn (mg/L)	0.29	0.34	0.32 *	0.1
Nitrate (mg/L)	1.67	1.74	1.7	20
Nitrite (mg/L)	0.06	1.63	0.85	3
pH	7.42	7.42	7.42	6.5–8.5
Biological
Total coliforms (CFU/100 mL)	55.4 × 106	4.5 × 106	30 × 106 *	0
E. coli (CFU/100 mL)	37.4 × 106	1.09 × 103	18.7 × 106 *	0

Notes: TDS, total dissolved solids; Fe, iron; Mn, manganese; NTU, Nephelometric Turbidity Unit; CFU, colony-forming unit; * exceeded quality standards. ¹ quality standards based on Ministry of Health Regulation Number 2, 2023.

shows the water quality parameters that exceeded quality standards between 2017 and 2019, together with their average concentrations (Table 3).

Between 2017 and 2019, the average well WPI in Jakarta was 5.9, which falls into the moderately polluted category. In 2017 and 2018, the well water quality at the majority of sample sites was categorized. However, in 2019 the majority of well water at the sample sites was classified as slightly polluted. Between 2017 and 2019, the number of wells in the province categorized as heavily polluted has decreased. More than 83% of the well water in Jakarta did not meet the physical, chemical, and biological water quality standards. The poorest well water quality was found in 2017, with 87.7% of the well water in the province not meeting standards (

Table 4. Frequency distribution of Well Water Pollution based on Water Quality Categories in Jakarta from 2017 to 2019.

Water Quality Category	2017		2018		2019
	Frequency	%	Frequency	%	Frequency	%
Interpretation of WIP Value
Good	32	12.3%	33	12.6%	43	16.5%
Slightly Polluted	73	28%	80	30.7%	115	44.1%
Moderately Polluted	112	42.9%	105	40.2%	78	29.9%
Heavily Polluted	44	16.9%	43	16.5%	25	9.6%
Total (N)	261	100%	261	100%	261	100%
Fulfillment of Standards
Meeting Standard	32	12.3%	33	12.6%	43	16.5%
Not Meeting Standard	229	87.7%	228	87.4%	218	83.5%
Total (N)	261	100%	261	100%	261	100%

Notes: IPj, pollution index value; good, IPj ≤ 1.0; slightly polluted, 1.0 < IPj ≤ 5.0; moderately polluted, 5.0 < IPj ≤ 10.0; heavily polluted, IPj > 10.0; IPj, pollution index value; meeting standard, 0 ≤ IPj ≤ 1.0; not meeting standard, IPj > 1.0.

).

Based on the well water quality map per subdistrict in Jakarta (Figure 1), it can be seen that well water quality in most subdistricts in the province was categorized as moderately polluted in 2017 and 2018, whereas in 2019, the majority of well water was categorized as lightly polluted. In 2017 and 2018, as shown by the water quality distribution map, heavily polluted wells were predominantly found in northeastern Jakarta. However, in 2019, there was a shift in the distribution pattern, with heavily polluted water occurring in northwestern Jakarta. The majority of wells with water quality categorized as good were located in South Jakarta.

3.2. Topographic, Sociodemographic, and Flood-Related Factors in Jakarta from 2017 to 2019

In terms of topographic factors, the average elevation in Jakarta is 15.4 m above sea level, which is classified as medium elevation. All of North Jakarta has an average elevation categorized as low, ranging from 2 to 5 m above sea level (masl), whereas the southern parts of East Jakarta and South Jakarta have higher average elevations, ranging from 42 to 63 m above sea level. The average rainfall in Jakarta between 2017 and 2019 was 175.64 mm, which is considered moderate rainfall. In terms of sociodemographic factors, the average population density between 2017 and 2019 was 23,730 people per square kilometer, categorized as a high population density. The average proportion of the population in Jakarta with a high education level was 61.22%, indicating a high level of education among residents. Regarding environmental factors, the average flood height in Jakarta between 2017 and 2019 was 34.03 cm, which is categorized as low flood severity (

Table 5. Distribution of Mean, Median, Standard Deviation (SD), and Minimum–Maximum for topographic, sociodemographic, and flood-related factors in Jakarta from 2017 to 2019.

Variables	Mean	Median	SD	Min-Max
Topographic
Elevation (masl *)	15.4	10	13.99	2.45–63.29
Rainfall (mm)	175.64	156	27.14	149–220
Sociodemographic
Population density (people/km2)	23,730	18,493	16,561	1196–95,676
Low education level (%)	38.78	37.43	7.21	17–69
High education level (%)	61.22	62.57	7.21	31–83
Environmental
Flood height (cm)	34.03	0	55.77	0–300

* Masl, meters above sea level.

).

From 2017 to 2019, more than 57% of subdistricts in Jakarta did not experience flooding. In terms of high and low flood levels in the subdistricts, the highest and lowest numbers were recorded in 2017 and 2019, respectively (5.5% and 26.1%, respectively) (

Table 6. Frequency of flood events, in addition to flood heights in Jakarta from 2017 to 2019.

Category	2017		2018		2019
	Frequency	%	Frequency	%	Frequency	%
No flood	150	57.5%	200	76.6%	173	66.3%
Low flood	64	24.5%	43	16.5%	68	26.1%
Medium flood	33	12.6%	14	5.4%	13	5.0%
High flood	14	5.4%	4	1.5%	7	2.7%
Total (N)	261	100%	261	100%	261	100%

Notes: low flood height, ≤100 cm; medium flood height, 101–150 cm; high flood height, >150 cm. Source: Tjasyono, 2012 [38].

).

3.3. Relationship Between Topographic, Sociodemographic, and Flood-Related Factors and Well Water Quality in Jakarta from 2017 to 2019

The Spearman correlation test results pointed to significance with weak relationships (|r| < 0.3) only between elevation and well water quality in Jakarta (

Table 7. Relationship between topographic, sociodemographic, and flood-related factors and well water quality in Jakarta from 2017 to 2019.

Variable	Water Pollution Index (WPI)
	Correlation Coefficient (r)	95% Confident Interval		R-Squared Coefficient (R2)
Topographic
Elevation	−0.27 **	−0.338 *	−0.201 *	0.037
Rainfall	−0.034	−0.106	0.038	0.00008024
Sociodemographic
Population density	0.087	0.016 *	0.15 *	0.003
Low education level (%)	0.078	0.01 *	0.14 *
High education level (%)	−0.078	−0.146 *	−0.104 *	0.002
Environmental
Flood height	−0.064	−0.132	0.003	0.008

* significant when the confidence interval does not include zero; ** significant weak at |r| < 0.3.

). For the elevation variable, the population correlation coefficient actually fell between −0.338 and −0.201 with a 95% confidence level. Since the entire confidence interval is below zero, this negative relationship indicates statistical significance [39]. Elevation and well water quality exhibited a weak negative correlation [40], indicating that the lower the elevation, the higher the WPI values, suggesting poorer well water quality, and vice versa.

For the rainfall variable, the correlation coefficient (r) of −0.034 indicates a very weak and almost nonexistent relationship between rainfall and the Water Pollution Index (WPI). The 95% confidence interval (−0.106, 0.038) includes zero, further confirming that the relationship is not significant.

For the population density variable, the correlation coefficient (r) of +0.087 indicates a very weak positive relationship between population density and the Water Pollution Index (WPI). The 95% confidence interval (0.016, 0.15) is entirely above zero, confirming that this positive relationship is significant, although the strength of the correlation was weak. According to the results, the higher the population density, the higher the WPI, indicating poorer well water quality, and vice versa.

For the low education variable, the correlation coefficient (r) of +0.078 indicates a very weak positive relationship between the percentage of low education and the Water Pollution Index (WPI). The 95% confidence interval (0.01, 0.14) is entirely above zero, confirming that this positive relationship is significant, although the strength of the correlation was weak. This finding indicated that the higher the proportion of residents with a low education level, the higher the WPI values, suggesting poorer well water quality, and vice versa.

For the flood variable, the correlation coefficient (r) of −0.034 indicates a very weak and almost nonexistent relationship between flood height and the Water Pollution Index (WPI). The 95% confidence interval (−0.132, 0.003) includes zero, further confirming that the relationship is not significant.

Thus, based on the correlation analysis results, it can be concluded that only the elevation variable exhibits a slightly stronger relationship (r = −0.27), although it is still considered practically weak according to the criteria of Bewick et al. [40]. Meanwhile, the other variables have effects that are too small to be deemed substantial for decision making or policy interventions related to water pollution (Water Pollution Index).

To assess the strength of the relationship between the Water Pollution Index (WPI) and various factors, including elevation, rainfall, population density, low education level, and flood height, this can also be observed through the scatter plot diagram (Figure 2). While all the variables showed weak correlations with WPI, elevation emerged as the variable with the strongest relationship. With an R² value of 0.037, elevation explained 3.7% of the variation in WPI, which, although still considered a weak correlation, was notably higher compared to the other variables. The fit line in the scatter plot indicates a negative correlation between elevation and WPI, suggesting that as elevation increases, WPI tends to decrease, indicating better water quality. The scatter points are somewhat spread along the line, but they also show some dispersion, indicating that while elevation affects WPI, other factors may contribute to the variation in water quality.

For population density, the fit line shows a positive correlation, meaning that as population density increases, WPI tends to increase, suggesting poorer water quality in more densely populated areas. However, the R² value of 0.003 indicates that this relationship is very weak. Similarly, for low education levels, the fit line also indicates a positive correlation, where higher percentages of people with low education levels are associated with higher WPI values, indicating worse water quality. With an R² of 0.002, the strength of this relationship is also very weak. The scatter points show some degree of clustering around the trend line, but there is still considerable variation in the data, reflecting the weak relationship between low education levels and WPI.

Rainfall and flood height had almost negligible effects, with R² values of 0.00008 and 0.008, respectively. The scatter points for both variables are widely spread out, with no clear trend, confirming that there is little to no correlation between rainfall or flood height and WPI. Despite these low R² values across the board, elevation remains the most significant factor in explaining variations in WPI, suggesting that geographical factors, particularly elevation, have a more substantial impact on water quality compared to the other variables.

The spatial analysis results, as shown by the overlay of the elevation and well water quality data (Figure 3), indicated that the lower the elevation, the higher the WPI, leading to poorer well water quality, and vice versa. As shown by the map, most of the subdistricts with heavily polluted well water are located in North Jakarta, which has the lowest elevation, between 2 and 5 masl. Conversely, South Jakarta, where the well water was predominantly categorized as slightly polluted or good, is situated at an elevation above 16 masl. However, when viewed in terms of trends from 2017 to 2019, no significant trend relationship is observed because there is heavily polluted water at all altitudes. Furthermore, wells with good water quality in 2017 were found to have transitioned to heavily polluted water in 2019 despite being in the same high-elevation location.

The spatial analysis of the overlay map of population density and well water quality data (Figure 4) indicated that as population density increased, the WPI also increased, leading to poorer well water quality. The map shows that over the three-year period, subdistricts categorized as having slightly polluted or good well water quality are predominantly found in South Jakarta, which has a relatively low population density. Most subdistricts in Central Jakarta, West Jakarta, and East Jakarta that were categorized as having moderately polluted or heavily polluted well water have high population densities. However, when observed in terms of trends from 2017 to 2019, no significant trend relationship is evident because there is heavily polluted water in all the areas regardless of whether the population density is low or high. Furthermore, wells with good water quality in 2017 were found to have transitioned to heavily polluted water in 2019 despite being in the same location within a medium-density population area. Additionally, the population density has not changed significantly over time, which also contributes to the absence of a noticeable trend relationship.

The spatial analysis of the overlay map of education level and well water quality data (Figure 5) indicated that as the proportion of residents with low education increased, the WPI also increased, leading to poorer well water quality. Conversely, as the proportion of residents with higher education increased, the WPI decreased, resulting in better well water quality. The map shows that over the three-year period, the majority of subdistricts with slightly polluted or good well water were located in South Jakarta, which has a higher proportion of residents with a high level of education. Subdistricts with heavily polluted and moderately polluted well water were in Central Jakarta, West Jakarta, and East Jakarta, where a higher proportion of residents have a low versus a high level of education. However, when observed in terms of trends from 2017 to 2019, no significant trend relationship is evident because there is heavily polluted water in all the areas regardless of whether the proportion of low education is high or low. Furthermore, wells with good water quality in 2017 were found to have transitioned to heavily polluted water in 2019 despite being in the same location with the same proportion of low education. Additionally, the proportion of low or high education has not changed significantly over time, which also contributes to the absence of a noticeable trend relationship.

The spatial analysis results of the overlay map of flood height and well water quality (Figure 6) revealed no relationship between these parameters in any of the subdistricts. The map indicates that over the three years, the majority of subdistricts categorized as having slightly polluted or good well water were located in the southern part of the province, which experienced moderate-to-high flood levels during the study period. In contrast, the subdistricts with heavily polluted and moderately polluted well water were located in Central Jakarta, West Jakarta, and East Jakarta and experienced either low flood levels or no flooding. Pejagalan subdistrict, which had the highest WPI in 2018, experienced no flooding during the three-year study period. Spatially, no significant trend relationship is evident, as heavily polluted water is present in all the areas regardless of whether they are in low or high flood elevation zones. Furthermore, wells that exhibited good water quality in 2017 were found to have transitioned to heavily polluted water in 2019 despite being in the same location with the same flood elevation. Additionally, flood events, whether in low or high flood elevation areas, have not changed significantly over time, further contributing to the absence of a discernible trend relationship.

4. Discussion

During 2017–2019, over 83% of the households in Jakarta continued to rely on poor-quality groundwater that failed to meet the hygiene and sanitation needs. The poor quality of groundwater in Jakarta from 2017 to 2019 can be attributed to several factors, including inadequate domestic wastewater management facilities in the province [41] and low education levels of the residents, which led to a lack of awareness regarding the importance of improving basic sanitation and consuming safe drinking water [5]. Areas with low topography, such as North Jakarta, are particularly vulnerable to pollution due to seawater intrusion and flooding [42].

The results revealed a significant with weak correlation between elevation and WPI value, meaning that lower elevations correspond to higher WPI values and poorer groundwater quality. This could be due to the tendency of contaminants from higher elevations to accumulate in lower areas. Additionally, pollutants tend to persist longer in the soil in lower-lying areas, where groundwater movement is slower compared to higher regions [43]. Furthermore, excessive groundwater extraction can lead to land subsidence. This subsidence, in turn, results in seawater Intrusion, which further exacerbates groundwater pollution, making the water quality worse [42]. Despite this, the correlation between elevation and WPI is weak. When observed through a spatial map trend, there is no significant relationship between the two variables. This could be influenced by hydrological factors, such as the interaction between rivers and groundwater, as well as the high chemical variability of groundwater at the local scale. This suggests that well water quality may be more influenced by specific local factors, such as direct pollution from septic tanks, industrial activities, or saltwater intrusion rather than general variables such as elevation or rainfall.

The results revealed no significant relationship between rainfall and WPI in Jakarta. These findings are consistent with research in Iran, which also found no significant correlation between rainfall and WPI [44]. Although long-term climate changes, such as rainfall patterns, can affect water quality, the influence of such changes may not always be evident in the short term. Thus, water quality at specific times may not reflect the direct impacts of changes in rainfall, as other factors may also influence water quality during that period [44].

The lack of significance of rainfall for WPI may also be attributed to the homogeneity of the secondary rainfall data used in this study. Secondary data affect the significance of correlation tests. For example, the rainfall data only distinguished between two different areas within the e province. The rainfall figures for North Jakarta, Central Jakarta, and West Jakarta were similar, resulting in minimal variation. In addition, there were discrepancies in the average rainfall data obtained from NASA and BMKG due to different measurement methods. In 2019, BMKG reported rainfall in West Jakarta of 173 mm [45], whereas NASA reported only 156 mm. Therefore, in future research, it will be essential to select data sources that more accurately reflect rainfall conditions in the region.

The analysis of sociodemographic factors, such as population density and education level, revealed a significant very weak correlation with WPI statistically, but when observed through a spatial map trend, there is no significant relationship between these variables. This could be due to the limited water management system in Jakarta, where the coverage of the wastewater disposal network serves only a small portion of the population. As a result, many households rely on septic tanks that often leak or discharge waste directly into surface water, contaminating groundwater. This leads to localized pollution, which may not always correlate with broader factors such as population density or education level [46].

The results revealed no significant relationship between flood events, and WPI, both statistically and in terms of spatial trends. This may be due to the ability of soil and rock conditions to act as natural filters, along with relatively deep aquifer systems that are protected from surface contamination, resulting in groundwater being less susceptible to pollution during floods. Furthermore, the city’s drainage system does not effectively separate stormwater from domestic and industrial wastewater [46]. During floods, contaminants from various sources are dispersed across different areas, resulting in heterogeneous pollution patterns. This explains why wells at the same elevation displayed good water quality in 2017 but were heavily polluted in 2019 regardless of flood height. In addition, government and local agency regulations regarding water resource management, infrastructure development, and community education on preserving water resources likely all play a role in improving groundwater quality and reducing the impact of floods on water resources [47].

Several factors may explain the lack of significance of floods in WPI. One possibility is that other factors, such as industrial activities and poor sanitation conditions, have a stronger impact on groundwater contamination. Additionally, discrepancies in data reporting may contribute to bias. For example, reports from the Jakarta Disaster Management Agency differ from findings in the literature, such as a 2018 report of a dike breach in Cilincing, which resulted in two days of flooding near residential areas [48]. Similarly, while a report indicated that the Sunter Jaya subdistrict experienced flooding in 2018 with water levels reaching 10–30 cm [49], the Jakarta Disaster Management Agency’s records did not report any flooding in this area.

Another limitation lies in the timing and spatial resolution of the data. The alignment between flood events and water quality measurements may not have been precise, affecting the results. Additionally, flood data are recorded at the neighborhood unit (Rukun Warga, RW) level, whereas well water quality data are available for only one point per subdistrict, leading to potential inconsistencies in measurement locations and periods. Moreover, water quality fluctuations are influenced by temporal variations in pollution sources. Since our study spans multiple years, other factors such as land use changes, infrastructure development, and government interventions may have altered the contamination levels [46]. This complexity may hinder the identification of direct relationships between the variables under examination and the water quality index (WQI).

Despite these limitations, this study serves as a preliminary investigation into the factors influencing the groundwater quality in Jakarta. Future research should explore additional factors not examined in this study, particularly due to data limitations, such as the lack of sanitation data at the subdistrict level. Moreover, challenges such as incomplete records, measurement inconsistencies, reporting errors, and delays in documentation highlight the need for more comprehensive data collection and validation.

The Water Pollution Index (WPI) was chosen in this study due to its standardized methodology outlined in the Ministry of Environment and Forestry Regulation Number 27 of 2021 [35]. WPI integrates multiple physical, chemical, and biological parameters into a single pollution index, providing a straightforward assessment of water quality. However, compared to other indices such as the Canadian Water Quality Index (CWQI) and the National Sanitation Foundation Water Quality Index (NSFWQI), WPI has certain limitations. CWQI applies a flexible aggregation method that allows the customized weightings of different parameters based on specific environmental conditions, making it more adaptable for different water bodies and pollution sources [50]. Meanwhile, NSFWQI includes a broader set of indicators, incorporating public health-related parameters such as dissolved oxygen, fecal coliform, and biochemical oxygen demand, making it particularly useful, especially for drinking water assessments [51]. Despite their advantages, both CWQI and NSFWQI require more complex calculations and higher data availability, which may limit their applicability in regions with constrained monitoring resources [52]. By comparing these indices, future studies can explore how different methodologies influence water quality assessments and policy recommendations in Jakarta.

The findings of this study provide valuable insights for policymakers in Jakarta and other urban areas facing similar challenges, emphasizing the need for targeted interventions. Expanding piped water networks is crucial to reducing reliance on contaminated well water given the correlation between poor water quality and high population density. Enhancing public education campaigns on water safety and sanitation can help mitigate contamination risks, as awareness is strongly linked to education levels. Integrating groundwater management with urban planning, particularly in low-elevation areas prone to pollution, is essential for long-term sustainability. Additionally, strengthening water quality monitoring systems through systematic assessments and data-driven decision making is necessary to prevent further groundwater degradation. By highlighting these policy implications, our study contributes to the broader discourse on sustainable groundwater management in urban environments.

Authors should discuss the results and how they can be interpreted from the perspective of previous studies and of the working hypotheses. The findings and their implications should be discussed in the broadest context possible. Future research directions may also be highlighted to ensure a more robust understanding of groundwater quality dynamics in Jakarta.

5. Conclusions

From 2017 to 2019, the groundwater quality in most regions of Jakarta (>83%) did not meet the required standards. The groundwater quality was assessed based on nine water quality parameters according to the quality standards outlined in the Ministry of Health Regulation Number 2 of 2023. The analysis revealed an average WPI value of 5.9, classified as moderately polluted. The contributing factors included topography and rainfall, which affected the quality of groundwater and the susceptibility of water resources to pollution. The average elevation in Jakarta is 15.4 masl, indicating a low elevation, and the average rainfall over the three years was 175.64 mm, which is considered moderate. In terms of the sociodemographic factors, the population density of Jakarta was categorized as high (23,730 people/km²), and 61.22% of the population was found to have a high level of education. In terms of environmental factors, the average flood height in the province was 34.04 cm, categorized as low flooding, with 57% of Jakarta not experiencing flooding.

This study still has limitations related to incomplete data and uncertainties related to data validity and accuracy due to potential monitoring errors, mismatch of measurement locations, recording errors, and documentation delays. The use of primary data sources and other methodologies in analyzing the data may strengthen the results of the analysis of the relationship between variables. However, the results of the correlation tests and spatial analysis showed that elevation was significantly related to groundwater quality in Jakarta from 2017 to 2019.

To mitigate groundwater contamination, The Jakarta Environmental Agency should optimize groundwater monitoring, implement educational programs, and enhance public awareness to prevent contamination. Strengthening sanitation infrastructure and wastewater treatment is essential for compliance with water quality standards. Collaboration with the private sector can expand piped water networks, ensuring safe drinking water access. Additionally, the Provincial Disaster Management Agency should improve flood mitigation efforts and conduct training to minimize groundwater pollution risks.

Future research could explore the causal relationships of the variables examined in this study using different study designs, such as cross-sectional or case–control. Researchers could also investigate the impact of other potentially influential variables, such as sanitation, socioeconomic factors, industrial activities, groundwater levels, and domestic wastewater management systems on water quality, using updated data. Beyond examining the factors affecting groundwater quality, subsequent research could also investigate the impacts of poor groundwater quality on human health.