Comparing predicted against measured steroid estrogen concentrations and the associated risk in two United Kingdom river catchments.

Predicted concentrations of estrone, 17β-estradiol, and 17α-ethinylestradiol generated from a geographical information systems-based model (LF2000-WQX) have previously been used to assess the risk of causing intersex in male fish in the rivers of England and Wales, United Kingdom. Few measured data of sufficient quality and spatial extent have been available to verify this risk assessment. New measured data have been collected from sewage treatment plant effluents and the receiving waters upstream and downstream of these discharges from the Erewash River and the Avon River systems in England. The model results for these rivers were in good agreement with the measured values in terms of estradiol equivalents. Critically, the risk assessment based on the measured data gave a risk assessment nearly identical to that derived from the modeled results. For individual estrogens, 17α-ethinylestradiol was modeled best and estrone worst. Poor simulations reflected poor estimates of the effluent concentrations, which were more variable from day to day and between works of nominally similar type than is assumed in the model. In support of this, model results for the Erewash River, calculated using observed effluent concentrations, were in excellent agreement with the measured data. The model has proved to be adequate in predicting overall estrogenic potency, and therefore risk, along these rivers; however, improvements are possible, particularly in predicting STP removal efficiency and therefore effluent concentrations.