The National Oceanic and Atmospheric Administration’s (NOAA) National Water Model 3.0 (NWM 3.0) is an advanced hydrologic model that produces hydrologic guidance for around 4000 NWS river forecast locations, as well as millions of ungauged rivers that lack traditional river forecasts. However, NWM 3.0 exhibits considerable variability in flood forecasting performance spatially, especially when tasked with providing hydrologic guidance for smaller stream basins. In an effort to improve upon these inaccuracies, NOAA-OWP’s NextGen Hydrological Modeling Framework provides a model-agnostic and open-source framework that fosters community engagement and integration of different hydrological models. Through the use of CIROH’s NextGen In A Box (NGIAB) on the JetStream2’s High Performance Computing platform, this project leverages the NextGen Hydro Framework to run simulations for two small, flashy stream basins in Westfield, PA, and Elkland, PA. We specifically compare the modeled outputs from the NextGen-based NOAH-OWP and Conceptual Functional Equivalent (CFE) to the National Weather Service’s SNOW-17 and Sacramento Soil Moisture Accounting (SAC-SMA) run via the Community Hydrologic Prediction System (CHPS) and the NWM 3.0 Retrospective. Through the use of hydrologically relevant metrics such as the Nash-Sutcliffe Efficiency (NSE), Kling-Gupta Efficiency (KGE), and seasonal bias statistics our analysis highlights the advantages of each model in these local examples of small stream basins.

Results from the comparison show that both calibrated NextGen and CHPS runs improved greatly upon respective uncalibrated runs for both Westfield and Elkland and that calibrated CHPS marginally outperforms calibrated NextGen runs. Additionally, both lumped models outperformed the NWM 3.0 Retrospective in Westfield (there is no NWM 3.0 Retrospective run for Elkland). The insights gained through these model comparisons serve the purpose of informing future model selection and calibration methods in efforts to provide more timely and accurate flash flood warnings.