Estimating Discharge and Nonpoint Source Nitrate Loading to Streams From Three End‐Member Pathways Using High‐Frequency Water Quality Data

• Published in: Water resources research 2017-12, Vol.53 (12), p.10201-10216
• Main Authors: Miller, Matthew P., Tesoriero, Anthony J., Hood, Krista, Terziotti, Silvia, Wolock, David M.
• Format: Article
• Language: English
• Subjects: Agricultural runoff; Biogeochemistry; Conductance; Costs; Cybersecurity; Data; Data processing; Dilution; Discharge estimation; Drinking water; Ecological conditions; Ecosystems; Environmental regulations; Groundwater; Groundwater runoff; Growing season; Hydrology; Load; Mineral nutrients; Mitigation; Nitrates; Nitrogen; Nutrient loading; Nutrients; Pollution sources; Resistance; Rivers; Runoff; Solutes; Stream discharge; Stream flow; Streams; Surface water; Surface-groundwater relations; Vulnerability; Water pollution; Water quality; Water quality measurements; Water treatment; Watersheds
• ISSN: 0043-1397; 1944-7973
• DOI: 10.1002/2017WR021654

The myriad hydrologic and biogeochemical processes taking place in watersheds occurring across space and time are integrated and reflected in the quantity and quality of water in streams and rivers. Collection of high‐frequency water quality data with sensors in surface waters provides new opportunities to disentangle these processes and quantify sources and transport of water and solutes in the coupled groundwater‐surface water system. A new approach for separating the streamflow hydrograph into three components was developed and coupled with high‐frequency nitrate data to estimate time‐variable nitrate loads from chemically dilute quick flow, chemically concentrated quick flow, and slowflow groundwater end‐member pathways for periods of up to 2 years in a groundwater‐dominated and a quick‐flow‐dominated stream in central Wisconsin, using only streamflow and in‐stream water quality data. The dilute and concentrated quick flow end‐members were distinguished using high‐frequency specific conductance data. Results indicate that dilute quick flow contributed less than 5% of the nitrate load at both sites, whereas 89 ± 8% of the nitrate load at the groundwater‐dominated stream was from slowflow groundwater, and 84 ± 25% of the nitrate load at the quick‐flow‐dominated stream was from concentrated quick flow. Concentrated quick flow nitrate concentrations varied seasonally at both sites, with peak concentrations in the winter that were 2–3 times greater than minimum concentrations during the growing season. Application of this approach provides an opportunity to assess stream vulnerability to nonpoint source nitrate loading and expected stream responses to current or changing conditions and practices in watersheds. The input of nutrients, such as nitrogen, to streams has been identified as contributing to the classification of the ecological condition of 42% of wadeable streams in the United States as “poor.” Elevated nutrients increase the costs of drinking water treatment, and reduction of nutrient loading is necessary for compliance with environmental regulations in many watersheds. Effective mitigation of the impacts of nutrients on stream ecosystems requires an understanding of where nutrients are coming from; for example, in runoff from farm fields or in groundwater. However, when nutrients are delivered to streams from many different parts of a watershed it is difficult to determine how much of the nutrient load measured in a stream is from a given locati