Construction of concentration prediction model of 2-MIB caused by benthic cyanobacteria utilizing automatic measuring device

In recent years, the difficulties in water utilization related to musty odor which is produced and emitted by benthic cyanobacteria (Phormidium autumnale) has become obvious. Tama River, which is a water resource of Tokyo, has a clear water quality. However, high concentration 2-methylisoborneol (2-MIB) exceeding 200 ng/L produced by P. autumnale was detected from raw water of Ozaku purification plant. As a result, Ozaku purification plant carried out strict powdered activated carbon treatment in order to deal with 2-MIB. In this study, we aimed to construct a model equation for the prediction of the short-time transition and the maximum concentration of 2-MIB in order to utilize it for future maintenance and management.By applying the cross-correlation correlogram and multiple regression analysis to water quality data of fiscal year 2015 exceeding 3,000 derived from automatic measuring device in increments of one hour, we constructed a multiple regression model equation for predicting 2-MIB concentration. When a model equation was verified using the water quality data of fiscal year 2016, a high correlation coefficient of 0.853 between the predicted value and the measured value was obtained. As a result, we enable the precise prediction of 2-MIB concentration.