Operational, hyper-resolution hydrologic modeling over the contiguous U.S. using themulti-scale, multi-physics WRF-Hydro Modeling and Data Assimilation System.

Abstract:

Operational flood, flash flood and water supply forecasting is typically conducted using a host of different
observational and modeling tools that range widely in process complexity, spatial resolution and
observational data sources. While such tailored approaches can provide significant skill in specific water
forecasting applications, the lack of a more coordinated general approach can result in inconsistency between various forecast products and can inhibit transfer of information, methodologies between forecast systems. With the aim of improving the timeliness, consistency and spatial fidelity hydrologic prediction products, the U.S. National Weather Service has initiated an effort to provide street-level, water prediction services for the nation. This effort seeks to incorporate advances in hydrometeorological observing capabilities, new hydrologic data assimilation methodologies, improvements in hydrographic and geospatial information and advances in the ulitizion of high performance computers for process-based hydrologic modeling. This talk will summarize the proposed Initial Operating Capability (IOC) for national water prediction using the community WRF-Hydro modeling system, scheduled for operational execution during late spring of 2016. Four different configurations of the WRF-Hydro system are planned including an Analysis and Data Assimilation configuration, Short Range (0-2 day) and Medium Range (0-10 day) deterministic configurations and a Long Range (0-30 day) enesmble configuration. Streamflow analyses and forecasts from each model configurations will be produced on 2.7 million river reaches of the NHDPlusv2 hydrographic dataset. This presentation summarizes results from a number of different model development and benchmarking activities conducted as part of the IOC effort. Results from prototype real-time forecasting activities conducted during the 2015 National Flood Interoperability Experiment (NFIE) will be presented as will retrospective prediction experiment results from each of the IOC WRF-Hydro configurations. Emphasis will be placed on describing the expected model performance across the nation and in identifyng critical sources of error and prediction uncertainty as a means for prioritizing future research and development efforts.