A comprehensive global dataset of river nutrient concentrations, loads and streamflow

Access to high quality river water quality data is crucial for environmental research and informing policymakers about management decisions. Currently, publicly available river data is spread across several large national and international databases and sparse in both spatial and temporal coverage in much of the world. In response to this, we produced three related data outputs which we have combined into a single comprehensive dataset: (1) a global dataset of total nitrogen, phosphorus and streamflow measurements compiled from seventeen river databases, containing over around 20,000 unique monitoring sites for total nitrogen, 40,000 sites for total phosphorus and 32,000 streamflow sites, (2) a set of 1,292 total nitrogen sites and 2,302 phosphorus sites with daily water quality values from a concentration estimation model and (3) global estimates of nutrient loads and concentrations produced by extrapolating the second output to areas with low data coverage using random forests. The dataset is intended to aid in future hydrological research and can also be used to advise on local river management, identify hotspots which may be prone to high nutrient loads, and compute global nutrient fluxes form land to sea.