Developing an empirical relationship between rainfall characteristics and microplastic emissions in an urban river

Abstract

Rainfall runoff transports microplastics into urban rivers, but concentration responses vary between events. This study analysed a secondary dataset from the Takano River, Kyoto, Japan, to evaluate river flow as a predictor of microplastic concentration. Records for particles between 10 and 300 micrometres were examined using time series, correlations and regression. Light rainfall produced an early concentration peak, moderate rainfall showed coincident flow and concentration peaks, and heavy rainfall produced delayed peaks during flow recession. A power-law model fitted to the light and moderate events explained 84 percent of the variation in log-transformed particle concentration. Leave-one-out cross-validation gave a coefficient of determination of 0.81. River flow therefore provided a useful predictor within the observed event domain. However, the model estimates total concentration at the monitoring site without separating upstream inputs from local rainfall wash-off. Transfer to other catchments requires validation because land use, surface resistance and microplastic sources can alter transport. Heavy rainfall requires consideration of storage, remobilisation and delayed delivery.