Fig. 4.

Hybrid-model predictions of DO_B forced by observed atmospheric and TP_lake and simulations under alternative scenarios. (A) The time series of hybrid-model-predicted DO_B under historical conditions (purple) is compared to observed DO_B (gray) and previous parametric efforts (orange) for reference. (B) Hybrid-model simulations are shown under alternative scenarios of atmospheric temperature and TP_lake. The model is forced by the detrended historical time series, shifted by 0 °C (green), 1 °C (blue), and 3 °C (red) warming and a constant value of TP (TP_lake). These simulations constitute a critical quantitative capability for management, linking physical model predictions of winter mixing and EDM predictions of summer DO depletion to examine the percentage of time that deep-water DO dips below 4 mg⋅L⁻¹. Under 3 °C, the hybrid model estimates that the lake will be hypoxic the majority of the time, regardless of phosphorus levels. The less extreme (but increasingly unlikely) temperature scenarios suggest that the reoligotrophic lake state with the lowest TP loadings may not be optimal for lake health. Without broadening management actions (e.g., on the food web), further TP decreases are not predicted to benefit the lake.