Lognormal distributions capture site-specific variability in enteric virus concentrations in wastewater

Abstract

As more data on virus concentrations in influent water from wastewater treatment plants (WWTPs) becomes available, establishing best practices for virus measurements, monitoring, and statistical modelling can improve the understanding of virus concentration distributions in wastewater. To support this, we assessed the temporal variability of norovirus, adenovirus, enterovirus, and rotavirus concentrations in influent water across multiple WWTPs in Switzerland, the USA, and Japan. Our findings demonstrate that the lognormal distribution accurately describes temporal variations in concentrations for all viruses at all sites, outperforming the gamma and Weibull distributions, which fail to capture high variability. However, notable differences in variability and uncertainty were observed across systems, underscoring the need for site-specific assessments. Using lognormal parameters, we identified optimal monitoring frequencies that balance cost-effectiveness and precision. For most sites, weekly monitoring was sufficient to estimate the annual average concentration of enteric viruses within a 95% confidence interval of 0.5 log10. We further examined the mechanistic basis of the lognormal distribution, highlighting processes that drive its prevalence and shape the behavior of its upper tail. By integrating these insights, this study provides a statistical foundation for optimizing virus monitoring frameworks and informing public health interventions targeting wastewater systems.