Streamflow prediction is a long-standing hydrologic problem. Development of models for streamflow prediction often requires incorporation of catchment physical descriptors to characterize the associated complex hydrological processes. Across different scales of catchments, these physical descriptors also allow models to extrapolate hydrologic information from one catchment to others, a process referred to as “regionalization”. Recently, in gauged basin scenarios, deep learning models have been shown to achieve state of the art regionalization performance by building a global hydrologic model. These models predict streamflow given catchment physical descriptors and weather forcing data. However, these physical descriptors are by their nature uncertain, sometimes incomplete, or even unavailable in certain cases, which limits the applicability of this approach. In this paper, we show that by assigning a vector of random values as a surrogate for catchment physical descriptors, we can achieve robust regionalization performance under a gauged prediction scenario. Our results show that the deep learning model using our proposed random vector approach achieves a predictive performance comparable to that of the model using actual physical descriptors. The random vector approach yields robust performance under different data sparsity scenarios and deep learning model selections. Furthermore, based on the use of random vectors, high-dimensional characterization identifies the uniqueness of catchments, thereby improving regionalization performance in gauged basin scenario when physical descriptors are uncertain, or insufficient.

The Solver of Hydrological Unstructured Domain (SHUD) is an integrated multi-process, multi-scale, multi-timestep hydrologic model where the dominant hydrological processes are fully coupled using the semi-discrete Finite Volume Method. The hydrologic processes in land surface, aquifer and river are fully coupled and solved together. The high spatial and temporal resolution in SHUD provides detailed and reliable hydrologic metrics in a watershed. The SHUD System consists of SHUD (the hydrologic model) and SHUDtoolbox (a data processing tool kit). The new SHUD system provides capabilities for public data downloads, pre-processing, hydrologic modeling, automatic calibration, post-processing and spatial visualization, and is fully open-source and ready for hydrological modelers to use. Here we introduce the philosophy of SHUD, from perceptual to computational structures of watershed hydrology, and select two watersheds (one in Africa and the other in California) as examples to demonstrate the workflow and capabilities of SHUD System. In modeling these examples, we exploit national/global public datasets and exemplify the data management, hydrologic analysis, model calibration and visualization capabilities. This modeling system not only supports quick deployment of hydrologic modeling functionality, but also benefits community modeling in other hydrology-related research fields, such as limnology, agriculture, climate change and Coupled-Natural-Human systems.