Environmental Monitoring and Assessment· 2026Q2
Ranking microbial risk of self-supplied groundwater in urban Indonesia
- 0citations
- Q2SCImago
- 2026year
Short summary
A new method ranks microbial contamination risk in urban Indonesian groundwater using secondary data, identifying dug-wells (2.9x higher E. coli odds vs. boreholes) and population density as key risk factors.
AI-generated from the title and abstract; the full text is not read.
Key points
- A new model ranks urban Indonesian groundwater contamination risk using secondary data alone.
- Dug-wells have 2.9 times greater odds of E. coli contamination (>100 CFU/100 mL) than boreholes.
- Contamination risk significantly increases with population density, aquifer lithology, and rainfall.
- The model's predictive accuracy (AUC 0.715) is acceptable but systematically underpredicts high-risk areas.
AI-generated from the title and abstract; the full text is not read.
Abstract
Abstract In Indonesia, 61% of urban households use self-supplied groundwater, with 24% relying on it as a primary drinking water source. Many self-supplied sources contain faecal contamination, posing health risks. Identifying locations with elevated contamination risk can help target piped-water investments and interim safety measures, but collecting sufficient water quality data is costly. This study develops and validates a method to rank contamination risk across Indonesian cities using secondary data alone. Contamination probability was modelled using logistic regression applied to water quality data from 619 households in Indonesia’s 2020 national water quality survey. A composite city-level risk score was constructed by combining predicted contamination probability with the proportion of households relying on self-supplied groundwater, derived from Indonesia’s 2022 national socioeconomic survey, enabling risk ranking across all 99 cities. Dug-wells had 2.9 times greater odds of Escherichia coli >100 CFU/100 mL than boreholes ( p < 0.001), and contamination risk increased significantly with population density ( p = 0.006). Aquifer lithology and rainfall were additional predictors ( p < 0.1). Model discrimination was acceptable (AUC 0.715, 95% CI 0.667–0.764). Independent validation against four datasets (14 locations) revealed acceptable agreement but systematic underprediction of E. coli exceedance probability at four locations (root mean square error of 0.116 and mean bias of −0.067), indicating the model is more reliable for identifying lower-risk than higher-risk cities. Including groundwater depth and sanitary condition data could increase model reliability. This approach provides a cost-effective screening tool for prioritising water-safety interventions where urban self-supply is widespread and data is limited.
The authors' abstract, as published at the source. Environmental Monitoring and Assessment, 2026 · DOI ↗
The rest is in the Pofolia app
Takeaways and questions to the paper; new summaries every day for your field. Free.
Sign in on the web to openField: Nutrition and Dietetics
Nutrition and DieteticsNursing