Authors: Ziyu Li, Andy Wood, Daniel McKenzie, Jonathan M. Frame
Title: Training a differentiable CFE to estimate static parameters for use in NextGen
Presentation Type: Poster Presentation
Abstract: Hydrologic model calibration can be challenging even with sufficient observations to constrain local model parameters, and far more difficult when estimating parameters across large domains – a process known as parameter regionalization. In recent years, the use of machine learning in differentiable hydrologic modeling has shown potential to address this regionalization problem. Here, a neural network (NN) learns to predict model parameters from meteorological forcings and catchment attributes by optimizing its weights to reproduce observed streamflow. We investigate whether this approach can be used to determine static parameters for NOAA’s Next Generation Water Resources Modeling Framework (NextGen), specifically for the Conceptual Functional Equivalent (CFE) model, by embedding a differentiable version (dCFE) into the NeuralHydrology (NH) platform, followed by extracting the trained NN to use in parameter regionalization across CONUS. We introduce two ways of estimating static parameters: one we call the oracle mode which represents the best achievable performance under ideal conditions and another we call the operational mode which we intend for practical use. We compare both to dynamic parameters obtained using the same workflow. This presentation describes this effort, including the validation of NH-dCFE to dCFE and CFE, successes in three modes of training, and the challenges encountered. We also offer recommendations on strategies to advance this parameter estimation approach in the future.