2007 Volume 13 Pages 219-224
Reduction and control of pollutant load from non-point sources is a key issue to improve the water condition in lakes. Monitoring of water quality in rivers for this purpose requires continuous measurement because the pollutant load from non-point sources is highly variable in a process of rain runoff. A measurement technique based on empirical relations between the signals from optical sensors and the results from water quality analysis has a possibility of practical measurement for this purpose (Liu et al; 2007). In this paper, the method is applied to seven rivers flowing into Lake Kasumigaura. The results show the wide and stable applicability of the method. In addition, an artificial neural network is used in trail for the process of regression analysis, and it reproduces the time series of COD, TN and TP very well.
