Salinization of stream water and groundwater at daily to decadal scales in a temperate climate DOI Creative Commons
Michelle D. Shattuck, Hannah M. Fazekas, Adam S. Wymore

et al.

Limnology and Oceanography Letters, Journal Year: 2023, Volume and Issue: 8(1), P. 131 - 140

Published: Jan. 9, 2023

Abstract Elevated salt concentrations in streams draining developed watersheds are well documented, but the effects of hydrologic variability and role groundwater surface water salinization poorly understood. To characterize these effects, we use long‐term data (12–19 yr) high‐frequency specific conductance (SPC) collected from 13 across New Hampshire, USA. Concentration–discharge ( C – Q ) relationships for chloride (Cl − derived SPC showed distinct seasonal variability. Diluting behavior was common, flushing occurred autumn winter, suggesting that both runoff contribute salts to streams. Long‐term show although extreme flood events initially reduced rural streams, recovered preflood conditions about a decade. Chronic Cl exceedances urban during all seasons. This research suggests variation stream flow, application deicing agents play freshwater salinization.

Language: Английский

Revisiting the Origins of the Power‐Law Analysis for the Assessment of Concentration‐Discharge Relationships DOI
Adam S. Wymore, William Larsen, Dustin W. Kincaid

et al.

Water Resources Research, Journal Year: 2023, Volume and Issue: 59(8)

Published: July 18, 2023

Abstract Concentration‐discharge ( C ‐ Q ) relationships are frequently used to understand the controls on material export from watersheds. These analyses often use a log‐log power‐law function = aQ b determine relationship between and . Use of in dates two seminal papers by Francis Hall (1970, https://doi.org/10.1029/WR006i003p00845 (1971, https://doi.org/10.1029/WR007i003p00591 ), where he compared six increasingly complex hydrological models, concluding had greatest explanatory power. Hall's conclusions, however, were based limited data set, with assumptions regarding water volume storage, simple model selection criteria. While is applied widely, it has not been rigorously tested evaluated over 50 years. We reexamined original models across time scales using 8 years high‐frequency weekly specific conductance performance more sophisticated we found analysis remains one best performing other performed equally as well including log‐linear functional form. Model was similar at sub‐daily scale but varied sampling method. More poorly relative simpler tended underpredict concentration flow extremes due constraints fitting parameters observed data. conclude, analyzed here, that suitable for analyses, opportunities exist refine differentiate among underlying distribution, recession applying reactive solutes.

Language: Английский

Citations

14
In-situ fluorescence spectroscopy indicates total bacterial abundance and dissolved organic carbon DOI Creative Commons
James Sorensen,

Mor Talla Diaw,

Abdoulaye Pouye

et al.

The Science of The Total Environment, Journal Year: 2020, Volume and Issue: 738, P. 139419 - 139419

Published: May 19, 2020

We explore in-situ fluorescence spectroscopy as an instantaneous indicator of total bacterial abundance and faecal contamination in drinking water. Eighty-four samples were collected outside the recharge season from groundwater-derived water sources Dakar, Senegal. Samples analysed for tryptophan-like (TLF) humic-like (HLF) in-situ, cells by flow cytometry, potential indicators such thermotolerant coliforms (TTCs), nitrate, a subset 22 samples, dissolved organic carbon (DOC). Significant single-predictor linear regression models demonstrated that most effective predictor TLF, followed on-site sanitation density; TTCs not significant predictor. An optimum multiple-predictor model TLF incorporated cells, nitrite, density, sulphate (r2 0.68). HLF was similarly related to same parameters with being best correlated (ρs 0.64). In sources, DOC clustered HLF, 0.84). The intergranular nature aquifer, timing study, and/or non-uniqueness signal can explain associations between TLF/HLF density nutrients but TTCs. population relates is likely be subsurface community develops based on availability matter originating sources. In-situ instantly indicates source impacted it remains unclear how specifically microbial risk this setting.

Language: Английский

Citations

37
Seawater nitrate assessment using a correction algorithm for temperature and pressure up to 100 MPa DOI
Xingyue Zhu, Wanzhao Cui, Naixin Zhang

et al.

Measurement, Journal Year: 2025, Volume and Issue: unknown, P. 117526 - 117526

Published: April 1, 2025

Language: Английский

Citations

0
Stuck at Home: Machine‐Learning Models Predicting Solute Concentrations of One Stream Failed to Predict Solute Concentrations in Other Streams DOI

Hollis C. Harrington,

Mark B. Green, John L. Campbell

et al.

Hydrological Processes, Journal Year: 2025, Volume and Issue: 39(5)

Published: May 1, 2025

ABSTRACT Machine‐learning models have been surprisingly successful at predicting stream solute concentrations, even for solutes without dedicated sensors. It would be extremely valuable if these could predict concentrations in streams beyond the one which they were trained. We assessed generalisability of random forest by training them or more and testing another. Models made using grab sample sensor data from 10 New Hampshire rivers. As observed previous studies, trained capable accurately that stream. However, on produced inaccurate predictions other streams, with exception measured sensors (i.e., nitrate dissolved organic carbon). Using multiple watersheds improved model results, but performance was still worse than mean dataset (Nash–Sutcliffe Efficiency < 0). Our results demonstrate machine‐learning thus far reliably only where trained, as differences concentration patterns sensor‐solute relationships limit their broader applicability.

Language: Английский

Citations

0
Predicting high‐frequency variation in stream solute concentrations with water quality sensors and machine learning DOI Creative Commons
Mark B. Green, Linda H. Pardo, Scott W. Bailey

et al.

Hydrological Processes, Journal Year: 2020, Volume and Issue: 35(1)

Published: Dec. 3, 2020

Abstract Stream solute monitoring has produced many insights into ecosystem and Earth system functions. Although new sensors have provided novel information about the fine‐scale temporal variation of some stream water solutes, we lack adequate sensor technology to gain same for other solutes. We used two machine learning algorithms – Support Vector Machine Random Forest predict concentrations at 15‐min resolution 10 which eight specific sensors. The were trained with data from intensive sensing manual sampling (weekly) four full years in a hydrologic reference within Hubbard Brook Experimental New Hampshire, USA. algorithm was slightly better predicting than (Nash‐Sutcliffe efficiencies ranged 0.35 0.78 compared 0.29 0.79 Machine). Solute predictions most sensitive removal fluorescent dissolved organic matter, pH conductance as independent variables both algorithms, least oxygen turbidity. predicted calcium monomeric aluminium estimate catchment yield, changed dramatically because it concentrates discharge. These results show great promise using combined approach discrete build high‐frequency solutes an appropriate or proxy is not available.

Language: Английский

Citations

29
Quantifying the frequency of synchronous carbon and nitrogen export to the river network DOI
Adam S. Wymore, Hannah M. Fazekas, William H. McDowell

et al.

Biogeochemistry, Journal Year: 2021, Volume and Issue: 152(1), P. 1 - 12

Published: Jan. 1, 2021

Language: Английский

Citations

26
Water Quality Sampling Frequency Analysis of Surface Freshwater: A Case Study on Bristol Floating Harbour DOI Creative Commons
Elisa Coraggio, Dawei Han, Claire Gronow

et al.

Frontiers in Sustainable Cities, Journal Year: 2022, Volume and Issue: 3

Published: Jan. 31, 2022

Water quality monitoring is essential to understanding the complex dynamics of water ecosystems, impact human infrastructure on them and ensure safe use resources for drinking, recreation transport. High frequency in-situ systems are being increasingly employed in schemes due their much finer temporal measurement scales possible reduced cost associated with manual sampling, manpower time needed process results compared traditional grab-sampling. Modelling data at higher reduces uncertainty allows capture transient events, although potential constraints storage, inducement noise, power conservation it worthwhile not using an excessively high sampling frequency. In this study, recorded Bristol's Floating Harbour as part local UKRIC Urban Observatory activities presented analyse events captured by current laboratory analysis scheme. The components time-series analysed work towards necessary temperature, dissolved oxygen (DO), fluorescent organic matter (fDOM), turbidity conductivity indicators quality. This study first its kind explore a statistical approach determining optimum different parameters dataset. Furthermore, provides practical tools understand how frequencies representative changes.

Language: Английский

Citations

17
Using In-Situ Optical Sensors to Understand the Biogeochemistry of Dissolved Organic Matter Across a Stream Network DOI Creative Commons
Adam S. Wymore, Jody D. Potter, B. Rodriguez-Cardona

et al.

Water Resources Research, Journal Year: 2018, Volume and Issue: 54(4), P. 2949 - 2958

Published: April 1, 2018

Abstract The advent of high‐frequency in situ optical sensors provides new opportunities to study the biogeochemistry dissolved organic matter (DOM) aquatic ecosystems. We used fDOM (fluorescent matter) examine spatial and temporal variability carbon (DOC) nitrogen (DON) across a heterogeneous stream network that varies concentration. Across ten streams explained twice concentration DOC ( r 2 = 0.82) compared DON 0.39), which suggests N‐rich fraction DOM is either more variable its sources or bioreactive than stable C‐rich fraction. Among sites, molar fluorescence was approximately 3x correlated with changes inorganic N, indicating both composition as well highly responsive N. Laboratory results also indicate we perform excitation‐emission wavelength pair generally referred “tryptophan‐like” peak when measured under laboratory conditions. However, since neither field sensor not measurements large percentage variation concentrations, challenges still remain for monitoring ambient pool nitrogen. Sensor networks provide insights into potential reactivity sites. These are needed build spatially explicit models describing dynamics water quality.

Language: Английский

Citations

30
Causes and Impacts of Decreasing Chlorophyll-a in Tibet Plateau Lakes during 1986–2021 Based on Landsat Image Inversion DOI Creative Commons

Shuyu Pang,

Liping Zhu, Chong Liu

et al.

Remote Sensing, Journal Year: 2023, Volume and Issue: 15(6), P. 1503 - 1503

Published: March 8, 2023

Lake chlorophyll-a (Chl-a) is one of the important components lake ecosystem. Numerous studies have analyzed Chl-a in ocean and inland water ecosystems under pressures from climate change anthropogenic activities. However, little research has been conducted on variations Tibet Plateau (TP) because its harsh environment limited opportunities for situ data monitoring. Here, we combined 95 measured concentration points Landsat reflection spectrum to establish an inversion model concentration. For this, retrieved mean annual past 35 years (1986–2021) 318 lakes with area > 10 km2 TP using backpropagation (BP) neural network prediction method. Meteorological hydrological data, quality parameters, glacier basin, along geographic information system (GIS) technology spatial statistical analysis, were used elucidate driving factors changes lakes. The results showed that displayed overall decrease during 1986–2021 (−0.03 μg/L/y), but 63%, 32%, 5% total number exhibited no significant change, decrease, increase, respectively. After a slight increase 1986–1995 (0.05 significantly decreased 1996–2004 (−0.18 μg/L/y). Further, it slightly 2005–2021 (−0.02 was negatively correlated precipitation (R2 = 0.48, p < 0.01), air temperature 0.31, surface (LSWT) 0.51, 0.42, volume 0.77, 0.01). non-glacial-meltwater-fed higher than those glacial-meltwater-fed lakes, except periods. Our shed light impacts variation lay foundation understanding

Language: Английский

Citations

8
Climate Variability Drives Watersheds Along a Transporter‐Transformer Continuum DOI
Hannah M. Fazekas, William H. McDowell, James B. Shanley

et al.

Geophysical Research Letters, Journal Year: 2021, Volume and Issue: 48(21)

Published: Nov. 4, 2021

Abstract We examined how climate variability affects the mobilization of material from six watersheds. analyzed one to seven years high‐frequency sensor data a temperate ecosystem and tropical rainforest. applied windowed analysis correlate concentration‐discharge (C‐Q) behavior with anomalies, providing insight into hydrological biogeochemical processes change in response variability. Positive precipitation anomalies homogenized C‐Q responses for dissolved organic matter, nitrate, specific conductance turbidity, indicating that dominate signal watersheds act as “conveyor belts” material. In contrast, drier warmer conditions led associated variation solute concentration, suggesting are primary control on export their flow. Results indicate can move along continuum transporter‐to‐transformer biologically active solutes potentially vary by biome.

Language: Английский

Citations

20