Author(s): Wang Manna; Yoshimura Chihiro; Kimura Fuminori
Linked Author(s):
Keywords: No Keywords
Abstract: For investigating the underlying drivers of odor production, we applied empirical dynamic modeling to the longterm water quality dataset of Kamafusa Reservoir in Japan. This application showed the nonlinear nature of the system, and allowed us to explore the causal relations as well we making odor prediction in Kamafusa Reservoir. Water temperature, pH, transparency, light intensity, and Green Phormidium were determined as causal variables of MIB production, and Green Phormidium was the most important variable for odor prediction. The modeling used in this study can be a powerful tool in causality identification and odor prediction, thus making contributions to reservoir management.
Year: 2018