Developing and refining ensemble streamflow forecasts in Nextgen and the RFC Hydrologic Ensemble Forecast Service

Objective:

Implement a NextGen testbed collection of research catchments for developing and benchmarking ensemble forecasting capabilities.

Abstract:

Over the past decade, the National Water Model (NWM) has developed limited ensemble
prediction products in the NWM; meanwhile the NWS River Forecast Centers (RFCs) have
deployed core ensemble prediction through the Hydrologic Ensemble Forecast Service (HEFS)
and the Met-Model Ensemble Forecast System (MMEFS). The new flexibility of Nextgen allows
for introducing full ensemble methods in the NWM and an opportunity to refine and
benchmark variations on HEFS methods. The overarching goal of this project is to connect NOAA
ensemble hydrologic forecasting science and development efforts with both the ‘next
generation’ Water Resources Modeling framework (Nextgen) as well as the RFC HEFS
capabilities, which is implemented in the Community Hydrologic Prediction System (CHPS), and
the supporting communities for each forecasting activity. This will be accomplished by
developing approaches to extend the Nextgen capabilities to incorporate ensemble forecast
methods and models, including those pioneered by HEFS, and to use Nextgen experimentation
to identify and test improved methods and datasets that can be used to advance HEFS. The
work would advance the state of the practice by creating state-of-the-art ensemble forecasting
capabilities in Nextgen, and ultimately in the NWM, while addressing previously identified gaps
in HEFS performance. The research will build from the foundation of the existing Office of
Water Prediction HEFS Testbed efforts, which have identified approximately 100 catchments
across the US and Alaska as a focus for research and development on HEFS techniques. It will
use hindcasts to intercompare performance of Nextgen and HEFS ensemble forecast
approaches, and seek to cross-fertilize high-performing methods between the two operational
systems. The project directly supports CIROH Research Theme 1 (Advancing Operational
Prediction), and will expand the hydrologic forecast Testbed being developed in the current
CIROH Testbed project that is also led by PI Wood. A major goal is to align NOAA flagship
hydrologic forecasting services with the recognized need for uncertainty quantification, and
benefit operations through improved ensemble forecasting for risk-based decision making in
HEFS, Nextgen, and ultimately the NWM.