Seasonal sediment and nutrient transport patterns

Predicting in-stream water quality constituents at the watershed scale using machine learning

Predicting in-stream water quality is necessary to support the decision-making process of protecting healthy waterbodies and restoring impaired ones. Data-driven modeling is an efficient technique that can be used to support such efforts. Our objective was to determine if in-stream concentrations of contaminants, nutrients-total phosphorus (TP) and total nitrogen (TN) -total suspended solids (TSS), dissolved oxygen (DO), and fecal coliform bacteria (FC) can be predicted satisfactorily using machine learning (ML) algorithms based on publicly available datasets. To achieve this objective, we evaluated four modeling scenarios, differing in terms of the required inputs (i.e., publicly available datasets (e.g., land-use/land cover)), antecedent conditions, and additional in-stream water quality observations (e.g., pH and turbidity). We implemented five ML algorithms-Support Vector Machines, Random Forest (RF), eXtreme Gradient Boost (XGB), ensemble RF-XGB, and Artificial Neural Network (ANN) -and demonstrated our modeling framework in an inland stream-Bullfrog Creek, located near Tampa, Florida. The results showed that, while including additional water quality drivers improved overall model performance for all target constituents, TP, TN, DO, and TSS could still be predicted satisfactorily using only publicly available datasets (Nash-Sutcliffe efficiency [NSE] > 0.75 and percent bias [PBIAS] < 10%), whereas FC could not (NSE < 0.49 and PBIAS >25%). Additionally, antecedent conditions slightly improved predictions and reduced the predictive uncertainty, particularly when paired with other water quality observations (6.9% increase in NSE for FC, and 2.7% for TP, TN, DO, and TSS). Also, comparable model performances of all water quality constituents in wet and dry seasons suggest minimal season-dependence of the predictions (<4% difference in NSE and < 10% difference in PBIAS). Our developed modeling framework is generic and can serve as a complementary tool for monitoring and predicting in-stream water quality constituents.

Simulation of fluvial sediment dynamics through strategic assessment of stream gaging data: A targeted watershed sediment loading analysis

Near-channel sediment loading (NCSL) is localized and episodic, making it difficult to accurately quantify its cumulative contribution to watershed sediment loading, let alone predict the effects from changes in river discharge due to climate change or land management practices. We developed a methodological framework, using commonly available stream gaging data, for estimating watershed-scale NCSL, a feature generally absent in most watershed models. The method utilizes a network of paired gages that bracket the incised river corridors of 15 tributaries to the Minnesota River, in which near-channel sources are often the dominant contributors of sediment loading. For each set of paired gages, we calculate NCSL as the difference between the upstream and downstream sediment loading minus the field contribution between the gages. NCSL generally increases with river discharge when it exceeds the observed threshold benchmark in the tributaries of Minnesota River Basin; accordingly, we developed a predictive model for quantifying NCSL using river discharge as the independent variable. This approach provides a predictive basis for evaluating the impacts on near-channel sediment supply from increases in runoff and river discharge. Application of this approach includes evaluation of watershed-scale conservation trade-offs, where benefits of landscape management practices, such as wetlands and reservoirs are measured in terms of reduction in downstream near-channel sediment loading in the incised river corridors.

Field scale nitrogen load in surface runoff: Impacts of management practices and changing climate

The use of Nitrogen (N) fertilizer boosted crop production to accommodate 7 billion people on Earth in the 20th century but with the consequence of exacerbating N losses from agricultural landscapes. Land management practices that can prevent high N load are constantly being sought for mitigation and conservation purposes. This study was aimed at evaluating the impacts of different land management practices under projected climate scenarios on surface runoff linked N load at the field scale level. A framework to analyze changes in N load at a high spatiotemporal resolution under high greenhouse emission climate projections was developed using the Pesticide Root Zone Model (PRZM) for the Willow Creek Watershed in the Fort Cobb Experimental Watershed in Oklahoma. Specifically, 12 combinations of land management and climate scenarios were evaluated based on their N load via surface runoff from 2020 to 2070. Results showed that crop rotation practices lowered both the N load and the probability of high N load events. Spring application reduced the negative effects in summer and fall from other land management practices but at the risk of increased probability of generating high N load in April and May. The fertilizer application rate was found to be the most critical factor that affected the amount and the probability of high N load events. By adopting a target application management approach, the monthly maximum N can be decreased by 13% while the annual mean N load by 6%. The model framework and analysis method developed in this research can be used to analyze tradeoffs between environmental welfare and economic benefits of N fertilizer at the field scale level.

Riparian erosion vulnerability model based on environmental features

Riparian erosion is one of the major causes of sediment and contaminant load to streams, degradation of riparian wildlife habitats, and land loss hazards. Land and soil management practices are implemented as conservation and restoration measures to mitigate the environmental problems brought about by riparian erosion. This, however, requires the identification of vulnerable areas to soil erosion. Because of the complex interactions between the different mechanisms that govern soil erosion and the inherent uncertainties involved in quantifying these processes, assessing erosion vulnerability at the watershed scale is challenging. The main objective of this study was to develop a methodology to identify areas along the riparian zone that are susceptible to erosion. The methodology was developed by integrating the physically-based watershed model MIKE-SHE, to simulate water movement, and a habitat suitability model, MaxEnt, to quantify the probability of presences of elevation changes (i.e., erosion) across the watershed. The presences of elevation changes were estimated based on two LiDAR-based elevation datasets taken in 2009 and 2012. The changes in elevation were grouped into four categories: low (0.5 - 0.7 m), medium (0.7 - 1.0 m), high (1.0 - 1.7 m) and very high (1.7 - 5.9 m), considering each category as a studied "species". The categories' locations were then used as "species location" map in MaxEnt. The environmental features used as constraints to the presence of erosion were land cover, soil, stream power index, overland flow, lateral inflow, and discharge. The modeling framework was evaluated in the Fort Cobb Reservoir Experimental watershed in southcentral Oklahoma. Results showed that the most vulnerable areas for erosion were located at the upper riparian zones of the Cobb and Lake sub-watersheds. The main waterways of these sub-watersheds were also found to be prone to streambank erosion. Approximatively 80% of the riparian zone (streambank included) has up to 30% probability to experience erosion greater than 1.0 m. By being able to identify the most vulnerable areas for stream and riparian sediment mobilization, conservation and management practices can be focused on areas needing the most attention and resources.

Evaluating the impacts of agricultural land management practices on water resources: A probabilistic hydrologic modeling approach

Evaluating the effectiveness of agricultural land management practices in minimizing environmental impacts using models is challenged by the presence of inherent uncertainties during the model development stage. One issue faced during the model development stage is the uncertainty involved in model parameterization. Using a single optimized set of parameters (one snapshot) to represent baseline conditions of the system limits the applicability and robustness of the model to properly represent future or alternative scenarios. The objective of this study was to develop a framework that facilitates model parameter selection while evaluating uncertainty to assess the impacts of land management practices at the watershed scale. The model framework was applied to the Lake Creek watershed located in southwestern Oklahoma, USA. A two-step probabilistic approach was implemented to parameterize the Agricultural Policy/Environmental eXtender (APEX) model using global uncertainty and sensitivity analysis to estimate the full spectrum of total monthly water yield (WYLD) and total monthly Nitrogen loads (N) in the watershed under different land management practices. Twenty-seven models were found to represent the baseline scenario in which uncertainty of up to 29% and 400% in WYLD and N, respectively, is plausible. Changing the land cover to pasture manifested the highest decrease in N to up to 30% for a full pasture coverage while changing to full winter wheat cover can increase the N up to 11%. The methodology developed in this study was able to quantify the full spectrum of system responses, the uncertainty associated with them, and the most important parameters that drive their variability. Results from this study can be used to develop strategic decisions on the risks and tradeoffs associated with different management alternatives that aim to increase productivity while also minimizing their environmental impacts.

Upper washita river experimental watersheds: meteorologic and soil climate measurement networks

Hydrologic, watershed, water resources, and climate-related research conducted by the USDA-ARS Grazinglands Research Laboratory (GRL) are rooted in events dating back to the 1930s. In 1960, the 2927-km Southern Great Plains Research Watershed (SGPRW) was established to study the effectiveness of USDA flood control and soil erosion prevention programs. The size of the SGPRW was scaled back in 1978, leaving only the 610-km Little Washita River Experimental Watershed (LWREW) to be used as an outdoor hydrologic research laboratory. Since 1978, the number of measurement sites and types of instruments used to collect meteorologic and soil climate data have changed on the LWREW. Moreover, a second research watershed, the 786-km Fort Cobb Reservoir Experimental Watershed (FCREW), was added in 2004 to the GRL's outdoor research laboratories to further study the effects of agricultural conservation practices on selected environmental endpoints. We describe the SGPREW, FCREW, and LWREW and the meteorologic measurement network (historic and present) deployed on them, provide descriptions of measurements, including information on accuracy and calibration, quality assurance measures (where known), and data archiving of the present network, give examples of data products and applications, and provide information for the public and research communities regarding access and availability of both the historic and recent data from these watersheds.

Long-term environmental research: the upper washita river experimental watersheds, oklahoma, USA

Water is central to life and earth processes, connecting physical, biological, chemical, ecological, and economic forces across the landscape. The vast scope of hydrologic sciences requires research efforts worldwide and across a wide range of disciplines. While hydrologic processes and scientific investigations related to sustainable agricultural systems are based on universal principles, research to understand processes and evaluate management practices is often site-specific to achieve a critical mass of expertise and research infrastructure to address spatially, temporally, and ecologically complex systems. In the face of dynamic climate, market, and policy environments, long-term research is required to understand and predict risks and possible outcomes of alternative scenarios. This special section describes the USDA-ARS's long-term research (1961 to present) in the Upper Washita River basin of Oklahoma. Data papers document datasets in detail (weather, hydrology, physiography, land cover, and sediment and nutrient water quality), and associated research papers present analyses based on those data. This living history of research is presented to engage collaborative scientists across institutions and disciplines to further explore complex, interactive processes and systems. Application of scientific understanding to resolve pressing challenges to agriculture while enhancing resilience of linked land and human systems will require complex research approaches. Research areas that this watershed research program continues to address include: resilience to current and future climate pressures; sources, fate, and transport of contaminants at a watershed scale; linked atmospheric-surface-subsurface hydrologic processes; high spatiotemporal resolution analyses of linked hydrologic processes; and multiple-objective decision making across linked farm to watershed scales.