Data and model integration for hydrologic applications

Abstract:

As research studies grow in size and in scope, the use of component integration frameworks becomes a more common modeling practice. These software systems provide mechanisms for sharing data among disparate computational modules during simulation runtime, to collectively solve a given problem. This functionality is particularly desirable when modeling cross-disciplinary systems such as climate science and watershed hydrology. Many hydrologic models rely on weather parameters, which may be observed in the field or estimated by other models, to drive simulation. While much of this data is readily available, the method by which it is integrated into coupled hydrologic simulations is not a trivial process, especially when dealing with large data stores. This work investigates modes of integration for large datasets within coupled hydrologic models. The approach is demonstrated by incorporating publicly available observation data as well as preprocessed datasets within a coupled modeling framework. We present this work and demonstrate how the methodology can be leveraged to simulate watershed-scale hydrologic systems.