Improving regional hydrological modeling in the Iberian Peninsula using the VIC model and optimized WRF/Noah-MP parameterizations

ABSTRACT Accurate hydrological modeling is essential for managing water resources under climate change, especially in regions like the Iberian Peninsula (IP), characterized by diverse hydroclimatic conditions. This study aims to enhance regional hydrological simulation performance across the IP by optimizing the configuration of the Noah Land Surface Model with multiparameterization options (Noah-MP), coupled with the Weather Research and Forecasting (WRF) model, and subsequently integrating outputs into the Variable Infiltration Capacity (VIC) hydrological model. A total of 78 WRF/Noah-MP experiments were evaluated to determine the optimal setup for accurately simulating hydrologically relevant variables, such as daily maximum and minimum temperatures and precipitation, during contrasting climatic years (dry 2005 and wet 2010). Results showed significant variability in model performance depending on specific Noah-MP parameterizations, particularly regarding vegetation dynamics, canopy stomatal resistance, runoff and surface exchange processes. Experiments employing the Ball-Berry canopy resistance model with dynamic vegetation, Monin-Obukhov surface drag exchange parameterization, and the TOPMODEL runoff scheme demonstrated superior performance, notably reducing temperature biases and better capturing spatial temperature patterns. Conversely, equilibrium water table-based runoff and original Noah surface exchange schemes exhibited less suitable performance and larger biases. VIC hydrological simulations, driven by optimized WRF/Noah-MP outputs (2005–2014), substantially improved monthly streamflow simulations across the IP’s headwater basins, confirmed by validation metrics including the Kling-Gupta Efficiency and the Kolmogorov-Smirnov test. Overall, 82% of the basins achieved satisfactory hydrological simulation performance, underscoring the robustness and added value of the proposed model configuration.