A do-it-yourself water quality sensor network to elucidate contaminant signatures and improve land management advice.

Quantifying contaminant loads to estuaries is essential for setting effective limits on resource use and safeguarding ecological values. Traditional monitoring programmes often rely on infrequent sampling, which can substantially underestimate loads and obscure key transport processes. While commercial high-frequency sampling stations are prohibitively expensive, open-source, do-it-yourself technologies now offer affordable alternatives for continuous monitoring. Here, we deployed ten low-cost, in situ monitoring stations equipped with research-grade sensors across an intensively farmed catchment draining to a sensitive estuarine environment in the Bay of Plenty, New Zealand. Using artificial neural network models trained on concurrent grab samples, we converted sensor measurements into reliable 15-min estimates of contaminant concentrations. High-frequency load calculations revealed nitrogen, phosphorus, and sediment exports to be 6–87% greater than estimates derived from traditional monthly sampling. Moreover, time-series outputs uncovered distinct sub-catchment contaminant mobilisation and transport dynamics that would otherwise remain undetected. These findings demonstrate that open-source, high-frequency monitoring can substantially improve contaminant load quantification, provide new insights into catchment processes, and inform the development of land management and policy strategies that reflect the unique spatio-temporal patterns of contaminant export.