The quest for the missing plastics: Large uncertainties in global river plastic export into the sea

Plastic pollution in the natural environment is causing increasing concern at both the local and global scale. Understanding the dispersion of plastic through the environment is of key importance for the effective implementation of preventive measures and cleanup strategies. Over the past few years, various models have been developed to estimate global river plastic export into the oceans, using limited plastic observations in river systems. However, there is a large discrepancy between the amount of plastic being modelled to leave the river systems, and the amount of plastic estimated to be in the ocean.  Here we perform a careful uncertainty analysis of the riverine plastic export estimates, examining both observational uncertainty, model parameter uncertainty, and underlying assumptions in models. Among the quantifiable sources of uncertainty, the conversion of visual plastic observations to plastic mass estimates introduces the largest uncertainty, leading up to three orders of magnitude uncertainty in the final mass estimates in most observations. Model structure and parameter uncertainty add an additional order of magnitude of uncertainty. Additionally, most global models assume that variations in the plastic flux are primarily driven by river discharge. However, we show that within the largest currently available datasets, correlations between river discharge (and other environmental drivers) and the plastic flux are never above 0.5, and strongly vary between catchments. Overall, we conclude that the yearly plastic load in individual rivers as well as the global riverine plastic flow into the ocean may be substantially less well-constrained than indicated by previous studies.