Search for: All records

Abstract Accurately estimating stream discharge is crucial for many ecological, biogeochemical, and hydrologic analyses. As of September 2022, The National Ecological Observatory Network (NEON) provided up to 5 years of continuous discharge estimates at 28 streams across the United States. NEON created rating curves at each site in a Bayesian framework, parameterized using hydraulic controls and manual measurements of discharge. Here we evaluate the reliability of these discharge estimates with three approaches. We (1) compared predicted to observed discharge, (2) compared predicted to observed stage, and (3) calculated the proportion of discharge estimates extrapolated beyond field measurements. We considered 1,523 site-months of continuous streamflow predictions published by NEON. Of these, 39% met our highest quality criteria, 11% fell into an intermediate classification, and 50% of site-months were classified as unreliable. We provided diagnostic metrics and categorical evaluations of continuous discharge and stage estimates by month for each site, enabling users to rapidly query for suitable NEON data. 

Abstract. Quantifying continuous discharge can be difficult, especially for nascent monitoring efforts, due to the challenges of establishing gauging locations, sensor protocols, and installations. Some continuous discharge series generated by the National Ecological Observatory Network (NEON) during its pre- and early-operational phases (2015–present) are marked by anomalies related to sensor drift, gauge movement, and incomplete rating curves. Here, we investigate the potential to estimate continuous discharge when discrete streamflow measurements are available at the site of interest. Using field-measured discharge as truth, we reconstructed continuous discharge for all 27 NEON stream gauges via linear regression on nearby donor gauges and/or prediction from neural networks trained on a large corpus of established gauge data. Reconstructions achieved median efficiencies of 0.83 (Nash–Sutcliffe, or NSE) and 0.81 (Kling–Gupta, or KGE) across all sites and improved KGE at 11 sites versus published data, with linear regression generally outperforming deep learning approaches due to the use of target site data for model fitting rather than evaluation only. Estimates from this analysis inform ∼199 site-months of missing data in the official record, and can be used jointly with NEON data to enhance the descriptive and predictive value of NEON's stream data products. We provide 5 min composite discharge series for each site that combine the best estimates across modeling approaches and NEON's published data. The success of this effort demonstrates the potential to establish “virtual gauges”, sites at which continuous streamflow can be accurately estimated from discrete measurements, by transferring information from nearby donor gauges and/or large collections of training data.