Need for Process Based Empirical Models for Water Quality Management: Salinity Management in the Delaware River Basin

Meyer, Eliot S.; Sheer, Daniel P.; Rush, Paul V.; Vogel, Richard M.; Billian, Hannah

doi:10.1061/(asce)wr.1943-5452.0001260

Cited by 6 publications

(4 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The ML model showed NSE = 0.85 during the testing period and NSE = 0.96 during the training period. Somewhat worse performance was reported by Meyer et al (2020) who used process-based and empirical models to simulate specific conductivity at several locations within the Delaware Bay estuary. They report NSE = 0.706-0.834 for the process-based empirical model and 0.458-0.744 for a hydrodynamic model.…”

Section: Model Resultsmentioning

confidence: 87%

Deep learning of estuary salinity dynamics is physically accurate at a fraction of hydrodynamic model computational cost

Gorski,

Cook,

Snyder

et al. 2024

Limnology & Oceanography

View full text Add to dashboard Cite

Salinity dynamics in the Delaware Bay estuary are a critical water quality concern as elevated salinity can damage infrastructure and threaten drinking water supplies. Current state‐of‐the‐art modeling approaches use hydrodynamic models, which can produce accurate results but are limited by significant computational costs. We developed a machine learning (ML) model to predict the 250 mg L−1 Cl− isochlor, also known as the “salt front,” using daily river discharge, meteorological drivers, and tidal water level data. We use the ML model to predict the location of the salt front, measured in river miles (RM) along the Delaware River, during the period 2001–2020, and we compare predictions of the ML model to the hydrodynamic Coupled Ocean–Atmosphere‐Wave‐Sediment Transport (COAWST) model. The ML model predicts the location of the salt front with greater accuracy (root mean squared error [RMSE] = 2.52 RM) than the COAWST model does (RMSE = 5.36); however, the ML model struggles to predict extreme events. Furthermore, we use functional performance and expected gradients, tools from information theory and explainable artificial intelligence, to show that the ML model learns physically realistic relationships between the salt front location and drivers (particularly discharge and tidal water level). These results demonstrate how an ML modeling approach can provide predictive and functional accuracy at a significantly reduced computational cost compared to process‐based models. In addition, these results provide support for using ML models in operational forecasting, scenario testing, management decisions, hindcasting, and resulting opportunities to understand past behavior and develop hypotheses.

show abstract

Section: Model Resultsmentioning

confidence: 87%

Deep learning of estuary salinity dynamics is physically accurate at a fraction of hydrodynamic model computational cost

Gorski,

Cook,

Snyder

et al. 2024

Limnology & Oceanography

View full text Add to dashboard Cite

show abstract

“…(2017), Meyer et al. (2020), and Morawietz et al. (2011), as well as our own analysis in Figures 3 and 7 below, our pp approach characterizes residuals using a log ratio model:

{\lambda }_{t}=\mathrm{ln}\left(\frac{{S}_{t}}{{O}_{t}}\right)=\mathrm{ln}\left({S}_{t}\right)-\mathrm{ln}\left({O}_{t}\right)

where

{S}_{t}

and

{O}_{t}

are the simulated and observed streamflows in time

t

, respectively.…”

Section: Stochastic Watershed Modeling Methodologymentioning

confidence: 99%

“…Aggregation of all sources of uncertainty into model error leads to some very attractive and simple pp approaches, resulting in a relatively transparent and straightforward generation of streamflow ensembles (Evin et al, 2013;Hunter et al, 2021;Koutsoyiannis & Montanari, 2022;Meyer et al, 2020;Sikorska et al, 2015;Zha et al, 2020). The fundamental challenge becomes selection and estimation of a suitable probabilistic model that can characterize the non-normality, heteroscedasticity, and very high stochastic persistence associated with DWM errors (see Hunter et al (2021) and McInerney et al (2017)), a choice which can have a tremendous impact on both DWM model parameter estimation and SWM prediction intervals (Evin et al, 2013).…”

Section: Post-processing Approaches For Generating Streamflow Ensembl...mentioning

confidence: 99%

Stochastic Watershed Model Ensembles for Long‐Range Planning: Verification and Validation

Shabestanipour

Brodeur

Farmer

et al. 2023

Water Resources Research

Self Cite

View full text Add to dashboard Cite

The introduction of stochastic streamflow models by Fiering (1967), Maass et al. (1962), and others led to a revolution in water resources planning, design, and management. These models enabled hydrologists to generate representative streamflow ensembles over future planning horizons, needed to explore the consequences of future hydrologic conditions not experienced historically, and formally characterize the reliability, vulnerability, and resilience of water resource systems (Hashimoto et al., 1982;Loucks & van Beek, 2017). Traditional stochastic streamflow models are typically statistical models rather than mechanistically driven hydrologic models. Such stochastic streamflow models may be adjusted to reflect changes in seasonality or other statistical properties of flow (e.g., Quinn et al., 2018), but tying statistical hydrologic changes to climate and land use change is

show abstract

“…Salinity is used as a longitudinal dispersion tracer in most water quality modelling exercises in estuaries [20]. Hydrodynamic models provide water quality models with important details, includes advection, dispersion, vertical mixing, temperature, and salinity [21]. A significant number of studies involved with the investigation of the spatial and temporal distribution of salinity in the estuary system [17], using a wide range of hydrodynamic models, including 1D [22], 2D [23,24], and 3D models [10,25].…”

Section: Water Salinity Modelmentioning

confidence: 99%

Impact of Flood Mitigation Project on the River Water Salinity

et al. 2022

View full text Add to dashboard Cite

The impact of flood mitigation project in the Kemaman River Basin was assessed in this study. Salinity intrusion was simulated in the study area by 1D numerical model. A 1-D hydrodynamic model coupled with a salinity model was used to analyze the salinity intrusion within Chukai River after the implementation of the flood mitigation project. The model was calibrated and validated using the data measured in Chukai River at 3 points from January 2007 until August 2013. Water quality simulation of salinity has been carried out once an excellent hydrodynamic model was established. The simulated river flow was reasonably matched to the measured data with R2 value 0.88, 0.92 and 0.82, respectively. Results suggest that after the realignment of Chukai River, the seawater intrudes further to the upstream river, causing the increasing salinity in the river about 10 - 15 ppt. However, with the floodway development, the channel would allow more water from Kemaman River being discharged into Chukai River. Increased in the volume of water in Chukai River has led to seawater dilution. Further, it invades the unique stretch of Chukai River and takes the salinity back to the initial state. Findings from the implementation of the flood mitigation project in the Kemaman river basin has benefitted the local society, watershed, and the surrounding biota ecosystems. Importantly, a greater prevention with the risk of repetitive flood damage to the buildings and structures in Kemaman area which has significantly achievable. HIGHLIGHTS Salinity model is used for flood mitigation project High salinity in Chukai river resulted from seawater intrusion Hydrodynamic model is to assess the water quality simulation

show abstract

Need for Process Based Empirical Models for Water Quality Management: Salinity Management in the Delaware River Basin

Cited by 6 publications

References 22 publications

Deep learning of estuary salinity dynamics is physically accurate at a fraction of hydrodynamic model computational cost

Deep learning of estuary salinity dynamics is physically accurate at a fraction of hydrodynamic model computational cost

Stochastic Watershed Model Ensembles for Long‐Range Planning: Verification and Validation

Impact of Flood Mitigation Project on the River Water Salinity

Contact Info

Product

Resources

About