Exploring the relationship between mobile phone data and wastewater flows: evidence from five Swiss catchments

Abstract

Emissions from urban drainage systems can have unwanted consequences for human and environmental health. Unfortunately, traditional water quality monitoring in sewers is expensive and not comprehensive enough to provide detailed data on pollution across an entire catchment. However, with the increasing digitization of society, alternative data sources such as mobile phone data offer new opportunities to assess wastewater production and dynamics. In this study, we investigate the relation between mobile phone data and wastewater flows in five catchments in Switzerland with different characteristics and sizes, using data from the largest Swiss telecom provider and simple multiple linear regression models. The initial results of this study are promising, although the degree of correlation observed between mobile phone data and wastewater production is rather low (max. R²=0.73) and varies greatly from catchment to catchment. As expected, we find non-linear effects in the data and more advanced modeling approaches, e.g. considering flow distances or dynamic wastewater travel times in the sewer network, may be needed to develop reliable predictions. In addition, we find that privacy protection issues currently limit the applicability in small catchments. Thus, we expect that the mobile data will benefit from domain-specific preprocessing for wastewater applications. The potential applications of this approach are far-reaching, including applications in urban drainage, wastewater treatment, drinking water, wastewater-based epidemiology and climate change adaptation. Last, but not least, wastewater flow or pollution data could even improve the data quality of the mobility data.