A software monitor for intermittent bacteria contamination in urban rivers

Norwegian receiving waters are of such high water quality that authorities consider opening them for bathing. The leading parameter to monitor the quality of bathing waters is fecal coliform bacteria (FC). For this parameter no rapid detection method is available. The main objective of this case study was to find a way to quickly predict bacteria contamination by observing different online parameters such as flow, conductivity or spectral absorption coefficient (SAC). In this study historical data from 1994 to 2000 was analyzed, and over a period of five weeks water samples were taken and analyzed for bacteria. The analysis of the historical data revealed fundamental sampling problems, which made the data useless for the purpose of this study. The analysis of the data collected for this study showed that exceeding the bathing water standard for bacteria can be predicted by evaluating the SAC with an acceptable accuracy. Furthermore a simple river quality model was implemented, including bacteria as a load fraction. With the help of rain data and discharge predictions expected bacteria numbers exceeding the bathing water standard could also be forecast.