A comparative analysis of national water model versions 2.1 and 3.0 reveals advances and challenges in streamflow predictions during storm events
Study RegionTexas, United States of America. Study FocusThis study presents a thorough evaluation of the National Water Model (NWM) versions 2.1 and 3.0 using an extensive dataset from 610 USGS gauges across Texas, and provides a detailed comparison of the model performance across the region. New Hy...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-04-01
|
| Series: | Journal of Hydrology: Regional Studies |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2214581825000205 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Study RegionTexas, United States of America. Study FocusThis study presents a thorough evaluation of the National Water Model (NWM) versions 2.1 and 3.0 using an extensive dataset from 610 USGS gauges across Texas, and provides a detailed comparison of the model performance across the region. New Hydrological InsightsThe National Water Model (NWM), launched by the National Oceanic and Atmospheric Administration (NOAA), represents a significant advancement in hydrological forecasting in the United States. However, the accuracy and reliability of NWM outputs, particularly in specific regions and during extreme events, require comprehensive evaluation. This study evaluates the accuracy and reliability of the NWM’s streamflow predictions in Texas by comparing them to observational data from USGS gauges. We assess the performance of NWM versions 2.1 and 3.0, focusing on both retrospective data and short-range forecasts. NWM version 3.0 achieves a 38.1% improvement in overall predictive accuracy over NWM 2.1, as measured by the Root Mean Square Error (RMSE). It also shows an 87.5% improvement in the Kling–Gupta Efficiency (KGE), reflecting better balance in bias, variability, and correlation; a 172.7% enhancement in streamflow predictive skill, as indicated by the Nash–Sutcliffe Efficiency (NSE); and a 10.6% increase in normalized predictive skill, as measured by the Normalized Nash–Sutcliffe Efficiency (NNSE). However, NWM 3.0 consistently underestimates streamflow and, in some cases, results in negative NSE and KGE values, highlighting difficulties in capturing extreme events. Short-range forecasts with data assimilation outperform retrospective analyses, highlighting the importance of data assimilation methods. Despite these advancements, challenges remain in predicting both the magnitude and timing of peak flow events underscoring the need for further improvements in the model to enhance its reliability. |
|---|---|
| ISSN: | 2214-5818 |