Numerical models are often used in hydrology to represent the complex, physical flow processes that occur in natural systems. Although simplified representations of reality, numerical models can be quite complex. This is especially true for integrated hydrologic models, which try to represent different processes and how they affect other parts of the system (e.g., crop-water demand and groundwater pumping; groundwater-surface-water interactions; climate change and ET; etc.).
The complexity of integrated models can make them difficult for non-modelers to evaluate, especially since terms like “sensitivity,” “calibration,” and “uncertainty” that come with complex equations are often used when they are presented. We define these and other general modeling terms and their methods for a broad audience using the Scott Valley Integrated Hydrologic Model (SVIHM) as an example.
This talk will demonstrate how disagreement between simulated results and observations can inform the conceptual model of the system, that hydrologic models that do not simulate crop-water demand are missing parameters that strongly influence calibration results, and that weakly coupled integrated models are a viable, computationally efficient approach to properly reproduce observed groundwater-surface-water interactions.