In situ sensors that collect high-frequency data are used increasingly to monitor aquatic environments. These sensors are prone to technical errors, resulting in unrecorded observations and/or anomalous values that are subsequently removed and create gaps in time series data. We present a framework based on generalized additive and auto-regressive models to recover these missing data. To mimic sporadically missing (i) single observations and (ii) periods of contiguous observations, we randomly removed (i) point data and (ii) day- and week-long sequences of data from a two-year time series of nitrate concentration data collected from Arikaree River, USA, where synoptically collected water temperature, turbidity, conductance, elevation, and dissolved oxygen data were available. In 72% of cases with missing point data, predicted values were within the sensor precision interval of the original value, although predictive ability declined when sequences of missing data occurred. Precision also depended on the availability of other water quality covariates. When covariates were available, even a sudden, event-based peak in nitrate concentration was reconstructed well. By providing a promising method for accurate prediction of missing data, the utility and confidence in summary statistics and statistical trends will increase, thereby assisting the effective monitoring and management of fresh waters and other at-risk ecosystems.
more »
« less
This content will become publicly available on June 30, 2024
Understanding links between water-quality variables and nitrate concentration in freshwater streams using high frequency sensor data
Real-time monitoring using
- NSF-PAR ID:
- 10494127
- Editor(s):
- Larsen, Stefano
- Publisher / Repository:
- PLOS One
- Date Published:
- Journal Name:
- PLOS ONE
- Volume:
- 18
- Issue:
- 6
- ISSN:
- 1932-6203
- Page Range / eLocation ID:
- e0287640
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
more » « less
Abstract Accurate riverine phosphorus concentration measurements are often critical to meet watershed management goals. Phosphorus monitoring programs often rely on proxy variables such as turbidity and discharge and have limited ability to accurately estimate concentrations of dissolved phosphorus fractions that are most bioavailable. Optical water quality sensors can make subhourly measurements and have been shown to reduce uncertainty in load estimates and reveal high‐frequency storm dynamics for nitrate and dissolved organic carbon. We evaluated the utility of in situ UV‐Visible spectrophotometers to predict total, dissolved, and soluble reactive phosphorus concentrations in streams draining agricultural, urban, and forested land use/land covers. We present the first statistically validated application of optical water quality sensors to demonstrate how sensors may perform in predicting phosphorus fraction concentrations through training set models. Total phosphorus predictions from UV‐Visible spectra were optimal when models were site‐specific, and the proportion of variance explained was generally as high as or higher than the results of other studies that rely only on discharge and turbidity. However, root mean square errors for total phosphorus models were relatively high compared to the median concentrations at each site. Models to predict dissolved and soluble reactive phosphorus concentrations explained a greater proportion of the variance than any other known proxy variable technique, and results varied by land use/land cover. Though accuracy limitations remain, this approach has potential to predict concurrent total, dissolved, and soluble reactive phosphorus concentrations at a high frequency for many applications in water quality research and management communities.
-
more » « less
Abstract High‐frequency in situ sensors have enabled researchers to measure solute concentrations at a time scale that captures the variability in stream discharge. We analyzed discrete samples and high‐frequency time series of solutes to characterize how nitrate (NO3−) and fluorescent dissolved organic matter (fDOM; a proxy for dissolved organic carbon) respond to changes in discharge at annual and intra‐annual timescales across a stream network in New Hampshire, USA. NO3−and fDOM exhibited highly variable concentration‐discharge (c‐Q) behavior at intra‐annual scales. Transport limitation, source limitation, and chemostatic behavior were observed to occur within and among years in all our study watersheds. Annual assessment of c‐Q misclassified streams 31% of the time, as the annual time step missed seasonal and event‐induced shifts in c‐Q dynamics. In some instances, anomalous events lasting less than 5% of the year determine the annual c‐Q behavior for a site. Catchment land use appeared to drive some of the variability among watersheds in c‐Q relationships and their temporal variability. Forested streams had highly variable NO3−c‐Q behavior and streams draining watersheds with more development had greater variability in fDOM c‐Q behavior. Sample frequency impacts how hydrologic systems are characterized and extrapolating c‐Q behavior from discrete samples alone can bias interpretations of c‐Q dynamics and our understanding of solute transport.
-
Freshwater salinization from anthropogenic activities threatens water quality and habitat suitability for many lakes and rivers in North America. Recognizing that salinization is a stress on freshwater environments globally, research on watershed salt transport is necessary for informed management strategies. Prior to this research, there were few studies that examined salt export regimes along a river–lake continuum to investigate the drivers, temporal dynamics, and modulators of freshwater salinization. Here, we use high-frequency in situ monitoring to assess specific conductance–discharge (cQ) relationships, chloride concentrations and fluxes, and the role of lakes in downstream salt transport. The Upper Yahara River Watershed in southern Wisconsin, USA, is a mixed urban and agricultural watershed where the lakes' chloride concentrations have risen from < 5 mg L−1 in the 1940s to > 50–80 mg L−1 in 2021. Our results suggest cQ behavior depends on land use, with urban areas exhibiting more frequent mobilization events during stormflow and agricultural areas exhibiting predominantly dilution dynamics. In addition, chloride loading is driven by hydrology and watershed size whereas concentrations and yields are a function of anthropogenic drivers like urbanization. We demonstrate how an in-network lake attenuates downstream salinity, dampening the hydrologic, anthropogenic, and seasonal patterns observed in rivers upstream of the lake. Importantly, biogeochemical processes in lakes overlay a seasonal signal on salinity that must be considered when investigating temporal dynamics of anthropogenic salinization. This research contributes to understanding of temporal dynamics of salt export through watersheds and can be used to inform management strategies for habitat protection.more » « less
-
Water quality is affected by multiple spatial and temporal factors, including the surrounding land characteristics, human activities, and antecedent precipitation amounts. However, identifying the relationships between water quality and spatially and temporally varying environmental variables with a machine learning technique in a heterogeneous urban landscape has been understudied. We explore how seasonal and variable precipitation amounts and other small-scale landscape variables affect E. coli, total suspended solids (TSS), nitrogen-nitrate, orthophosphate, lead, and zinc concentrations in Portland, Oregon, USA. Mann–Whitney tests were used to detect differences in water quality between seasons and COVID-19 periods. Spearman’s rank correlation analysis was used to identify the relationship between water quality and explanatory variables. A Random Forest (RF) model was used to predict water quality using antecedent precipitation amounts and landscape variables as inputs. The performance of RF was compared with that of ordinary least squares (OLS). Mann–Whitney tests identified statistically significant differences in all pollutant concentrations (except TSS) between the wet and dry seasons. Nitrate was the only pollutant to display statistically significant reductions in median concentrations (from 1.5 mg/L to 1.04 mg/L) during the COVID-19 lockdown period, likely associated with reduced traffic volumes. Spearman’s correlation analysis identified the highest correlation coefficients between one-day precipitation amounts and E. coli, lead, zinc, and TSS concentrations. Road length is positively associated with E. coli and zinc. The Random Forest (RF) model best predicts orthophosphate concentrations (R2 = 0.58), followed by TSS (R2 = 0.54) and nitrate (R2 = 0.46). E. coli was the most difficult to model and had the highest RMSE, MAE, and MAPE values. Overall, the Random Forest model outperformed OLS, as evaluated by RMSE, MAE, MAPE, and R2. The Random Forest was an effective approach to modeling pollutant concentrations using both categorical seasonal and COVID data along with continuous rain and landscape variables to predict water quality in urban streams. Implementing optimization techniques can further improve the model’s performance and allow researchers to use a machine learning approach for water quality modeling.more » « less
