Pollutant dispersion modelling for Portuguese river water uses protection linked to tracer dye experimental data

Mathematical models are well known as useful tools for water management practices, directly or indirectly related to the implementation of the Water Framework Directive (WFD) in European countries. They can be applied to solve or understand either simple water quality problems or complex water management problems of trans-boundary rivers or multiple-purpose and stratified reservoirs. Accidental spills of pollutants are of general concern and could be harmful to water users along the river, becoming crucial to get knowledge of the dispersive behaviour of such pollutants. In this context, the mathematical modelling of dispersion phenomena can play an important role. Additionally, a judicious selection of mathematical models for application in a specific river basin management plan can mitigate prediction uncertainty. Therefore, intervention measures and times will be established with better reliability and alarm systems could efficiently protect the aquatic ecosystems, the water uses and the public health. The main purpose of this paper is to evaluate the performance of river water systems dispersion modelling, based on tracer experiments data for calibration and validation. The present work describes the methodology used in the monitoring programs, basically consisting in the injection of a tracer dye (rhodamine WT) in an upstream river section and follow-up of the dye cloud along the river to determine the water dispersion behaviour in situ. The models were developed to simulate different water quality management scenarios on each reach of the three Portuguese rivers under study: Mondego, Douro and Tagus rivers. However, further developments are needed for Douro and Tagus rivers in order to simulate vertical and transversal dispersion processes and improve the model correlation with the experimental data. The models were calibrated and validated in order to produce operational tools used to estimate the probabilistic leading edge/peak/tail times, the pollutant losses by volatilization, adsorption, precipitation, etc. and remaining concentrations. These tools allows to define, for example, how long water intake need to be suspended after a pollutant spill and can be easily integrated in a future DSS, which should be developed and implemented by each one of the river basin management authorities. The good correlation between experimental and simulated data allows us to conclude that the applied models are accurate enough to describe and predict conservative pollutant transport under different hydrodynamic scenarios. This methodology is appropriate to assess the environmental impact of pollutant loads directly introduced into the streams and, subsequently, to define and implement the best water sources protection practices. Key-Words: Dispersion modelling; Water uses protection; Rhodamine; parameter estimation; Portuguese rivers.