Childhood lead exposure affects over 500,000 children under 6 years old in the US; however, only 14 states recommend regular universal blood screening. Several studies have reported on the use of predictive models to estimate lead exposure of individual children, albeit with limited success: lead exposure can vary greatly among individuals, individual data is not easily accessible, and models trained in one location do not always perform well in another. We report on a novel approach that uses machine learning to accurately predict elevated Blood Lead Levels (BLLs) in large groups of children, using aggregated data. To that end, we used publicly available zip code and city/town BLL data from the states of New York (n = 1642, excluding New York City) and Massachusetts (n = 352), respectively. Five machine learning models were used to predict childhood lead exposure by using socioeconomic, housing, and water quality predictive features. The best-performing model was a Random Forest, with a 10-fold cross validation ROC AUC score of 0.91 and 0.85 for the Massachusetts and New York datasets, respectively. The model was then tested with New York City data and the results compared to measured BLLs at a borough level. The model yielded predictions in excellent agreement with measured data: at a city level it predicted elevated BLL rates of 1.72% for the children in New York City, which is close to the measured value of 1.73%. Predictive models, such as the one presented here, have the potential to help identify geographical hotspots with significantly large occurrence of elevated lead blood levels in children so that limited resources may be deployed to those who are most at risk.
24 May 2022 in Journal of Colloid and Interface Science
The Mobile Lead Testing Unit (MLTU), coordinated by the Newark Water Coalition (NWC) and the University of California, Berkeley researchers, sought to measure and educate community members on the sources of lead exposure within the home by conducting field analysis on lead in paint, water, soil, and dust. Throughout our project spanning design, outreach, education and training, methodological design, analysis, and evaluations, the MLTU instilled, executed, and added to the principles of community-owned and managed research projects. The primary data collected will be used to build an exposure model and to support the NWC in their advocacy. Our community voice paper presents our reflections on the nuanced, unforeseen, and complexities of community-driven science and attempts to forge a path toward democratizing knowledge and science while fighting for environmental justice.
21 February 2025 in Environmental Research Letters
For decades, the dominant approach to lead poisoning has been to focus on homes affected by lead paint and to treat children who are already suffering from lead poisoning. This individualizing approach developed in the context of the defunding and deregulation of government agencies in the 1980s. In recent years, however, community–academic partnerships have reframed lead as an environmental issue produced by the development of the lead industry in the twentieth century and connected to overlapping histories of exploitation, discrimination, and inaction. These community-based projects have contributed to shifting research agendas (by emphasizing historical analysis and the study of the soil and dust), achieved policy changes (with a focus on community-level solutions), and built networks and solidarity with groups advocating for climate justice, tenant organizing, and food security.
Most North American drinking watermains were installed in the 1940s–1960s. A large share of these mains are well past their expected service life, with cast iron disproportionately driving the break rates. Most existing lifecycle costing frameworks, whether deterioration-based asset management models or standard MRR (Maintenance, Rehabilitation, and Replacement) cost tools, are tied to a fixed data level i.e., they are designed for data-rich utilities requiring asset and cost details most smaller utilities rarely hold. Additionally, many existing frameworks omit the effect of discounting which in turn obscures the NPV (Net Present Value) of future expenditure. This paper presents a data-tiered lifecycle cost framework, incorporating power-law non-homogeneous Poisson process break-rate modelling, risk-weighted prioritization, basic and advanced cost-assessment modules, and NPV discounting, to test whether proactive renewal strategies produce lower lifecycle costs than a fix-upon-break approach over 100 years and at what discount rate that advantage disappears. Applied to Kitchener (936 km) and Calgary (5,391 km), lining-heavy proactive strategies lower lifecycle cost by up to 12% and 20% at a 0% discount rate. The savings become negligible near 3% discount rate and reverse above it. Utilities can therefore run the framework on inventory and break records alone and establish whether proactive renewal remains defensible at their own discount rate.