Factors contributing to the minimum water column stability and timing of the winter turnover in the Ogouchi reservoir.

Turnover in lakes and reservoirs causes circulation in the water column from the bottom to the surface when the water column stability becomes low. Previous studies commonly mentioned that turnover occurs when stratification indices become small, but the threshold is rarely discussed. While turnover phenomena have been extensively studied by evaluating changes in bottom dissolved oxygen (DO), the relationship between the disappearance of hypoxia and water temperature indices has not been determined. This study focused on the factors influencing the minimum thermal gradient (TG) and Schmidt Stability Index (SSI), and the timing of turnover events using DO as an indicator of mixing in the Ogouchi reservoir from 1992 to 2001. The results showed that the occurrence of minimum TG and SSI is mainly driven by inflow retention time and average maximum wind speed. Moreover, minimum air temperature and outflow retention time have few contributions to minimum SSI. It was found that 7 out of 10 years exhibited full winter turnover, while the remaining years showed incomplete mixing with persistent hypoxia at the reservoir bottom. This study identifies four cases based on onset thresholds of 0.0035 °C m for TG and 30 J m for SSI to explain turnover event: Case 1: an ideal state with stratification indices below the threshold, resulting in the disappearance of hypoxia; Case 2: indices above the threshold sustain hypoxia; Case 3: an irregular state where the indices exceed the threshold, yet hypoxia disappears; and Case 4: an unexpected persistence of hypoxia despite being below the threshold. The majority of the years (70 percent) were explained by thresholds. The multiple regression analysis indicated the importance of wind speed on the turnover event. Therefore, the effect of wind shear was analyzed for 30 percent of the years that cannot be explained by thresholds (cases 3 and 4). Case 3 shows turnover occurrence due to strong accumulated wind shear, despite exceeding thresholds. Conversely, Case 4 reveals weak wind shear preventing bottom water upwelling, even below thresholds. In conclusion, the precise TG and SSI thresholds for the onset of turnover event were determined using DO data. The thresholds explained the occurrence and non-occurrence of turnover event in most of the years and wind speed clarified unexplained cases by thresholds. The presented method successfully evaluated the timing of turnover and can be applicable elsewhere.