Analysis and modeling of algal blooms in the Nakdong River, Korea

The purpose of this study is to improve the prediction accuracy of algal blooms in the Nakdong River, Korea, to support improved management practices. To improve algal bloom predictions, it is necessary to consider the microbiological characteristics of algal groups. Therefore, Korean governmental data, including water quality, flow rate, and meteorological data, over 12 years (2004–2015) were analyzed to characterize chlorophyll a (Chl-a) concentration dynamics in the Nakdong River, as the primary indicator of algae. The correlation between water temperature and Chl-a concentration differed by region at the study site. While a positive correlation (R = 0.63) was found in the relatively clean (i.e., mesotrophic) upstream area, a negative correlation (R = −0.51) was found in the more eutrophic downstream area. These results indicate that nutrients are a dominant factor of algal blooms in mesotrophic upstream areas, but other factors may have greater impacts in eutrophic downstream areas. Moreover, the dominance of different algal groups differed spatially and temporally at the study site. The three-dimensional hydrodynamics and water quality modeling-capable Environmental Fluid Dynamics Code model was applied to represent the hydrodynamics and kinetics of water quality variables, including Chl-a. Four statistics, the coefficient of determination (R2), Nash–Sutcliffe model efficiency (ME), percentage model bias (Pbias), and cost function (CF), were used to evaluate the model prediction accuracy against field observation data. Compared to the case in which only a single algal group was modeled, modeling multiple algal groups together improved the R2, ME, Pbias, and CF values from 0.25 to 0.50, 0.15 to 0.50, 30.43 to 4.98, and 0.56 to 0.48, respectively. The degree of improvement in the model prediction accuracy was greater for more eutrophic regions at the study site. These results show that there should be greater focus on studying multiple algal groups together when modeling algal bloom predictions to support the development of alternative management practices.