Zheng Duan

Understanding the spatiotemporal dynamics of dissolved organic carbon (DOC) is essential for predicting aquatic carbon fluxes and managing water quality in boreal catchments. Traditional DOC modelling approaches are limited in their ability to represent seasonal dynamics, event-driven responses, and riparian-zone contributions. Here, we evaluate the performance of the Krycklan-HYPE model in simulating streamflow and DOC concentrations across the Krycklan catchment in northern Sweden. The model was calibrated and validated for 14 sub-catchments using the Kling-Gupta Efficiency (KGE) metric. Results showed consistently good to very good performance for streamflow (calibration KGE = 0.50–0.82; validation KGE = 0.35–0.71), with low uncertainty and strong spatial transferability. In contrast, the success of DOC simulations was more varied, with KGE values across streams of 0.11–0.59 for the calibration period and 0.05–0.58 for the validation period (excluding one negative value). A negative correlation between streamflow simulation performance and DOC simulation performance was observed, indicating a trade-off between hydrological and biogeochemical simulation accuracy. Seasonal analysis showed that streamflow and DOC were best captured during spring and autumn, while DOC dynamics during summer and high-flow events were not well simulated. This pointed to the model structure limitation with missing representations of corresponding processes. Scenario-based analysis further highlighted that streamflow was best simulated under high flows, while DOC performance was better under low flow conditions. These results demonstrate the Krycklan-HYPE model’s strengths in streamflow simulation but also underscore limitations in representing DOC mobilization and transport. Future studies should focus on developing model structures that incorporate relevant processes, and adopting calibration strategies that account for spatially heterogeneous parameterization.