Event-scale hysteresis metrics to reveal processes and mechanisms controlling constituent export from watersheds: A review✰

Liu, Wenlong; Bírgand, François; Tian, Senlin; Cheng, Chen Li

doi:10.1016/j.watres.2021.117254

Cited by 28 publications

(30 citation statements)

References 87 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Future work could include using these sediment surrogate model outputs to test if the ML model is better at accounting for the processes controlling sediment transport as more accurate and representative features are calculated. The sediment surrogate model outputs could also be used to calculate a hysteresis‐index to quantify rising and falling limb hysteresis trends (Liu et al, 2021; Lloyd et al, 2016; Vaughan, Bowden, et al, 2017) to improve the ML models.…”

Section: Discussionmentioning

confidence: 99%

Using machine learning to improve predictions and provide insight into fluvial sediment transport

Lund

Groten

Karwan

et al. 2022

Hydrological Processes

View full text Add to dashboard Cite

A thorough understanding of fluvial sediment transport is critical to addressing many environmental concerns such as exacerbated flooding, degradation of aquatic habitat, excess nutrients, and the economic challenges of restoring aquatic systems. Fluvial sediment samples are integral for addressing these environmental concerns but cannot be collected at every river and time of interest. Therefore, to gain a better understanding for rivers where direct measurements have not been made, extreme gradient boosting machine learning (ML) models were developed and trained to predict suspended sediment and bedload from sampling data collected in Minnesota, United States (U.S.), by the U.S. Geological Survey. Approximately 400 watershed (full upstream area), catchment (nearby landscape), near-channel, channel, and streamflow features were retrieved or developed from multiple sources, reduced to approximately 30 uncorrelated features, and used in the final ML models. The results indicate suspended sediment and bedload ML models explain approximately 70% of the variance in the datasets. Important features used in the models were interpreted with Shapley additive explanation (SHAP) plots, which provided insight into sediment transport processes. The most important features in the models were developed to normalize streamflow by the 2-year recurrence interval and quantify the rate of change in streamflow (slope), which helped account for sediment hysteresis. Generally, this study also showed a combination of mostly watershed and catchment geospatial features were important in ML models that predict sediment transport from physical samples. This study is a promising step forward in making fluvial sediment transport predictions using machine learning models trained by physically collected samples. The approach developed here can be used wherever similar datasets exists and will be useful for landscape and water management.

show abstract

Section: Discussionmentioning

confidence: 99%

Using machine learning to improve predictions and provide insight into fluvial sediment transport

Lund

Groten

Karwan

et al. 2022

Hydrological Processes

View full text Add to dashboard Cite

show abstract

“…A correlation was further found between the climate parameters (temperature, evapotranspiration) and the soil hydraulic parameters (soil moisture, field water table). Other studies have found that temperature and precipitation can impact P movement (Liu et al 2021). The relationship between temperature, evapotranspiration and soil properties is reflected in the model.…”

Section: Correlation Between Model Parametersmentioning

confidence: 92%

Neural network model predictions for phosphorus management strategies on tile-drained organic soils

et al. 2022

View full text Add to dashboard Cite

The organic soils of Holland Marsh, Ontario are used for intensive vegetable production, which demands high-phosphorus (P) fertilizer applications. Such high-fertilizer applications on these tile-drained lands lead to eutrophication in surrounding water bodies. This study investigated the application of neural network (NN) models for deriving P management strategies. Seven NN models were assessed using the following two approaches: a time series with 1-year training and 1-year testing of the models and a randomization analysis where a random 80% of data were used for model training and the remainder for model testing. The feed-forward model using the randomization and the long-short-term memory model using time-series outperformed all other models. Two strategies for P management were evaluated: a direct approach that predicts P loads using new fertilizer rates or controlled drainage discharge rates and a particle swarm optimization (PSO) that used percent reduction of actual P loads to predict an optimal water table management strategy. Overall, the direct approach identified a water table level of 30 cm from the soil surface during the spring and 80 cm during the summer period as optimal to reduce P loads. The PSO analysis showed that a reduction of P loads by 20% in the spring and up to 40% in the summer through water table control would not compromise crop production.

show abstract

“…A novel aspect of the workflow is that it enables users to evaluate the impacts of different event delineations on c‐Q hysteresis metrics, which is an issue of growing interest to the c‐Q research community (Liu et al, 2021). In particular, the workflow described here covers steps 1–3 of the larger workflow described in Figure 3 of Liu et al (2021), a recently published review summarizing the methodologies of event‐scale c‐Q analyses. Below, we offer some brief concluding thoughts on important facets of the workflow, including key considerations for implementing the workflow, suggestions for modifying and improving the code, and ideas for future research in this area: It is important to note that this workflow is designed to be used with continuous, high‐frequency (sub‐daily) time series data sets of concentration and discharge.…”

Section: Concluding Thoughtsmentioning

confidence: 99%

“…of Liu et al (2021), a recently published review summarizing the methodologies of event-scale c-Q analyses. Below, we offer some brief concluding thoughts on important facets of the workflow, including key considerations for implementing the workflow, suggestions for modifying and improving the code, and ideas for future research in this area:• It is important to note that this workflow is designed to be used with continuous, high-frequency (sub-daily) time series data sets of…”

mentioning

confidence: 99%

An open‐source automated workflow to delineate storm events and evaluate concentration‐discharge relationships

Millar

Buda

Duncan

et al. 2022

Hydrological Processes

View full text Add to dashboard Cite

The advent of in situ optical sensors that can collect sub‐daily measurements of nutrients and turbidity in flowing water bodies has yielded comparatively much larger water quality data sets than were previously available. With these newly available data sets, there has been increased interest in studying event‐based concentration‐discharge (c‐Q) relationships to infer the sources and pathways of various watershed constituents during storms. With water quality data sets increasingly growing in size and scope, the need to automate the processing and analyses of such data has become apparent. However, consensus on storm event delineation methods as they pertain to c‐Q analysis is currently lacking, and methodological details, including parameter values, are sometimes unreported in the literature. Here, we present an open‐source workflow using the programming language R that automates the processing of sub‐daily c‐Q data to analyze event‐based hysteresis patterns. Briefly, the workflow accepts a time series of concentration and discharge data, extracts stormflow from streamflow, delineates storm events and then evaluates c‐Q relationships using widely applied metrics like the hysteresis index (HI) and the flushing index (FI). We applied the workflow to three watersheds in the mid‐Atlantic United States, including a 0.4‐km2 agricultural watershed, a 150‐km2 urbanizing watershed and a 29 940‐km2 mixed land use river basin. Sub‐daily sensor‐based nutrient concentrations and discharge data were collected in each watershed. Using the small agricultural watershed as an example, we demonstrate the step‐by‐step application of the workflow. We then present results from the larger watersheds as a means for comparison and to illustrate the flexibility of the code. We believe that this rapid approach to event‐based c‐Q analysis will allow scientists and practitioners more time to focus on interpreting results, and promote greater scientific reproducibility. Likewise, we conclude this Scientific Briefing with suggested future improvements to the workflow to increase the automation of data analyses and reproducibility.

show abstract

Event-scale hysteresis metrics to reveal processes and mechanisms controlling constituent export from watersheds: A review✰

Cited by 28 publications

References 87 publications

Using machine learning to improve predictions and provide insight into fluvial sediment transport

Using machine learning to improve predictions and provide insight into fluvial sediment transport

Neural network model predictions for phosphorus management strategies on tile-drained organic soils

An open‐source automated workflow to delineate storm events and evaluate concentration‐discharge relationships

Contact Info

Product

Resources

About