Qualitative Comparison of Streamflow Information Programs of the U.S. Geological Survey and Three Non-Federal Agencies

Norris, J. Michael; Lewis, Michael; Dorsey, M.E.; Kimbrough, Robert A.; Holmes, Robert R.; Staubitz, W.W.

doi:10.3133/ofr20071426

Cited by 3 publications

(2 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The United States Geological Survey (USGS) operates about 85% of all stream gages in US. According to USGS, average operation cost was about $14,000 annually per typical continuous stream gage (Norris et al., 2007) with most cost associated with site inspections and field experiments. Due to the cost limitation, most gages are deployed to large rivers, and the number of gages is rapidly decreasing over the past decades as a result of the shrinking budget allocation.…”

Section: Introductionmentioning

confidence: 99%

Application of Deep Learning for Imaging‐Based Stream Gaging

Boomen

Qian

2021

Water Resources Research

View full text Add to dashboard Cite

Stream gages track data vital for water resources management. Frequently collected streamflow data are invaluable to many hydrological analyses, such as the assessment of the magnitude and frequency of floods or droughts, evaluation of water quality for implementing the Total Maximum Daily Loads (TMDL). The United States Geological Survey (USGS) operates about 85% of all stream gages in US. According to USGS, average operation cost was about $14,000 annually per typical continuous stream gage (Norris et al., 2007) with most cost associated with site inspections and field experiments. Due to the cost limitation, most gages are deployed to large rivers, and the number of gages is rapidly decreasing over the past decades as a result of the shrinking budget allocation. Meanwhile the need for high quality streamflow data is ever increasing in order to meet the demand for better water resource management and resilience to climate change. For example, there are increasing demands for river stage data from smaller tributaries, which are very few but can greatly improve the performance of flood-forecasting models and warning systems (Kruger et al., 2016). To overcome these challenges, innovative solutions are demanded to substantially reduce the cost and hazards associated with current stream gauging operations.Despite great advances in flow measurements technologies, most USGS stream gages apply the identical method that has been used for more than 100 years. Specifically, discharge is not measured directly, it is calculated from the height of water surface, that is, the "stage," which can be measured more conveniently. Discharge is directly measured only periodically following the velocity-area method (Turnipseed & Sauer, 2010). An empirical relation between discharge and stage is established with repeated field experiments. The relation, or the "rating-curve," can then be applied to report discharge continuously based on stage measurements. Since it is often difficult, sometimes dangerous, to conduct field experiments during high flows, extrapolation of the rating curve is unavoidable to estimate discharge of extreme flood flows. Substantial errors can be introduced with rating curve extrapolation, which causes great uncertainties to flood frequency analysis (Lang et al., 2010;Le Coz et al., 2014).Non-contact or remote sensing solutions may solve difficulties of standard river gaging processes. Kruger et al. (2016) showed that bridge mounted ultrasonic sensors can be a low-cost alternative to traditional river gaging methods. It has been demonstrated that radars can be applied to detect both water depth and surface

show abstract

Section: Introductionmentioning

confidence: 99%

Application of Deep Learning for Imaging‐Based Stream Gaging

Boomen

Qian

2021

Water Resources Research

View full text Add to dashboard Cite

show abstract

“…We believe that statistical models need not be limited to small geographic areas in the same way, which is why we perform multi-basin flood prediction. Finally, current physics-based predictions from NOAA rely on stream measurements taken using the USGS network of roughly 11,000 stream gauges, which each cost between $7,000 and $15,000 per year to maintain [7]. If flood prediction can be done without reliance on these expensive in-situ measurements it could reduce costs for existing flood prediction systems as well as expand flood prediction efforts to areas that could not previously afford river gauging infrastructure.…”

Section: Introductionmentioning

confidence: 99%

Machine Learning for Generalizable Prediction of Flood Susceptibility

Sidrane,

Fitzpatrick,

Annex

et al. 2019

Preprint

View full text Add to dashboard Cite

Flooding is a destructive and dangerous hazard and climate change appears to be increasing the frequency of catastrophic flooding events around the world. Physicsbased flood models are costly to calibrate and are rarely generalizable across different river basins, as model outputs are sensitive to site-specific parameters and human-regulated infrastructure. In contrast, statistical models implicitly account for such factors through the data on which they are trained. Such models trained primarily from remotely-sensed Earth observation data could reduce the need for extensive in-situ measurements. In this work, we develop generalizable, multibasin models of river flooding susceptibility using geographically-distributed data from the USGS stream gauge network. Machine learning models are trained in a supervised framework to predict two measures of flood susceptibility from a mix of river basin attributes, impervious surface cover information derived from satellite imagery, and historical records of rainfall and stream height. We report prediction performance of multiple models using precision-recall curves, and compare with performance of naive baselines. This work on multi-basin flood prediction represents a step in the direction of making flood prediction accessible to all at-risk communities.

show abstract