Modeling Dissolved Organic Carbon Export from Water Supply Catchments in the Northeastern United States DOI
Kezhen Wang,

Rajith Mukundan,

Rakesh K. Gelda

et al.

Published: Jan. 1, 2024

Natural organic matter (NOM) in rivers is an important energy source to sustain aquatic ecosystem health. However, surface water supply systems where chlorination often used for disinfection, NOM also a precursor the carcinogenic and mutagenic disinfection byproducts such as trihalomethanes haloacetic acids. Effective management of maintain both functions high-quality requires better understanding transport patterns. operationally measured by dissolved carbon (DOC). Challenges using DOC data analysis on catchment scale largely relate spatial temporal variations DOC, low sampling frequency which fails capture multi-scale To help improve sources transport, we analyzed its long-term patterns six catchments New York City Water Supply System monitoring models. We tested empirical models prediction including linear, nonlinear time-series based model formulations. found that generalized additive (GAMs) produced most robust results across catchments. Then, applied calibrated GAM predict daily concentrations estimate fluxes analyze trends. Finally, compared relationships between features investigate regional differences, focusing mechanistic processes associated with parsing out hydrological signals. The showed hydrology plays larger role three top 5% streamflow corresponded nearly 50% annual export, whereas nutrient production were more others. study presents approach streams can inform targeted strategies waters.

Language: Английский

A Stacking Ensemble Model of Various Machine Learning Models for Daily Runoff Forecasting DOI Open Access

Mingshen Lu,

Qinyao Hou,

Shujing Qin

et al.

Water, Journal Year: 2023, Volume and Issue: 15(7), P. 1265 - 1265

Published: March 23, 2023

Improving the accuracy and stability of daily runoff prediction is crucial for effective water resource management flood control. This study proposed a novel stacking ensemble learning model based on attention mechanism prediction. The has two-layer structure with base meta model. Three machine models, namely random forest (RF), adaptive boosting (AdaBoost), extreme gradient (XGB) are used as models. to integrate output obtain predictions. applied predict inflow Fuchun River Reservoir in Qiantang basin. results show that outperforms models other terms accuracy. Compared XGB weighted averaging (WAE) 10.22% 8.54% increase Nash–Sutcliffe efficiency (NSE), an 18.52% 16.38% reduction root mean square error (RMSE), 28.17% 18.66% absolute (MAE), 4.54% 4.19% correlation coefficient (r). significantly simple indicated by both Friedman test Nemenyi test. Thus, can produce reasonable accurate reservoir inflow, which great strategic significance application value formulating rational allocation optimal operation resources improving breadth depth hydrological forecasting integrated services.

Language: Английский

Citations

52
Chemodiversity of dissolved organic matter exports from subtropical humid catchment driven by hydrological connectivity DOI
Xiao Hu, Yingtian Deng, Cong Zhou

et al.

Water Research, Journal Year: 2024, Volume and Issue: 260, P. 121902 - 121902

Published: June 7, 2024

Language: Английский

Citations

4
Modeling dissolved organic carbon export from water supply catchments in the northeastern United States DOI
Kezhen Wang,

Rajith Mukundan,

Rakesh K. Gelda

et al.

The Science of The Total Environment, Journal Year: 2025, Volume and Issue: 963, P. 178532 - 178532

Published: Jan. 20, 2025

Language: Английский

Citations

0
Identifying the determinants of the spatial patterns and temporal fluctuation characteristics of riverine pCO2 of the largest subtropical river using machine learning methods DOI Creative Commons

Menghan Chen,

Lei Cheng,

Li‐Wei Chang

et al.

Journal of Hydrology Regional Studies, Journal Year: 2025, Volume and Issue: 58, P. 102284 - 102284

Published: March 8, 2025

Language: Английский

Citations

0
Exploring hydrological controls on dissolved organic carbon export dynamics in a typical flash flood catchment using a process-based model DOI
Yue Wu, Hang Su, Lei Cheng

et al.

The Science of The Total Environment, Journal Year: 2024, Volume and Issue: 921, P. 171139 - 171139

Published: Feb. 23, 2024

Language: Английский

Citations

3
Contrasting variations of ecosystem gross primary productivity during flash droughts caused by competing water demand and supply DOI Creative Commons

Kaijie Zou,

Lei Cheng, Mengqi Wu

et al.

Environmental Research Letters, Journal Year: 2024, Volume and Issue: 19(2), P. 024031 - 024031

Published: Jan. 22, 2024

Abstract Flash drought events (FDEs) are projected to increase frequently in a warming world, significantly impacting ecosystem productivity and the global carbon cycle. The development of FDEs, induced by anomalies different environmental variables, may cause responses ecosystem’s gross primary (GPP). However, GPP variations underlying mechanisms during FDEs have rarely been quantified. This study collected long-term (>10 years) high-quality flux observations from FLUXNET 2015 dataset investigate their driving FDEs. Results showed that all vegetation types two contrasting One variation is decreasing then increasing standardized anomaly (V-shape response). other shows an followed (inverted V-shape response was increased soil water content deficit at onset stage In contrast, inverted net radiation Such results indicated competing moisture supply atmospheric demand controlling with its development. Moreover, contribution use efficiency magnitude (64.5 ± 22.4%) greater than (47.6 18.7%). identified across multiple which can improve our ability predict future effects more frequent on productivity.

Language: Английский

Citations

2
Modelling of sap flux density of oak in a humid region in China DOI
Yujie Liu, Zhengbing Chen, Lei Cheng

et al.

Ecohydrology, Journal Year: 2024, Volume and Issue: 17(5)

Published: April 23, 2024

Abstract Transpiration plays a vital role in determining the watershed water cycle. However, we still have little knowledge of characteristics tree transpiration Hanjiang River Basin, which is source for middle route South‐to‐North diversion project. Here, measured sap flux density oak trees ( Quercus , dominant species here) at 10‐min resolution 2 years and explored its response to environmental conditions. The incoming short‐wave radiation vapour pressure deficit well explained variation daytime density, statistical model was then proposed calculate correspondingly; nighttime module based on calculation. Sap showed clear counter‐clockwise hysteresis radiation, clockwise deficit. Our can reproduce corresponding This study unravelled controls an efficient simulation, provided important understanding forests' use, critical significance availability

Language: Английский

Citations

1
Multi-factor weighted image fusion method for high spatiotemporal tracking of reservoir drawdown area and its vegetation dynamics DOI Creative Commons

Shiqiong Li,

Lei Cheng,

Li‐Wei Chang

et al.

International Journal of Applied Earth Observation and Geoinformation, Journal Year: 2024, Volume and Issue: 129, P. 103855 - 103855

Published: April 25, 2024

Reservoir drawdown areas (RDAs) with distinct dry-wet cycles and vegetation dynamics have emerged as significant hotspots for carbon-related activities. However, high-resolution spatiotemporal tracking of the variations RDAs remains challenging because they often change dramatically are controlled by both human activities natural factors. Herein, a modified image fusion method was proposed to capture rapid in integrating impact factor information into analysis. The capability tested Danjiangkou (DJK) it is largest artificial freshwater lake Asia highly variable RDA, since surrounded gently sloping plains or hills. results showed that workflow produced reliable predictions (r=0.83,RMSE=0.097) compared original Enhanced Spatial Temporal Adaptive Reflectance Fusion Model (ESTARFM) (r=0.60,RMSE=0.195), demonstrating improved mapping water surface changes dynamics. Using method, 15-d RDA were derived from 2013 2022 30-m resolution. interannual maximum estimated be 278 km2 after dam elevated 2013. Normalized Difference Vegetation Index (NDVI) decreased inundation frequency (IF) increased. Mean NDVI growing season (May–October) 0.109 (17.6 %) 0.156 (26.0 under 30 %–40 % IF 60 %–70 IF, respectively, 0 %–10 which referred "natural" considering its rare inundation. Moreover, mean length only 63 19 d respectively. Furthermore, 77.3 exhibited decrease NDVI, whereas 22.7 an unusual increase, possibly due selection dominant species well-adapted during succession. Overall, this study not new high monitoring RDAs, but also highlighted importance within accurate estimation their carbon budgets.

Language: Английский

Citations

1
Modeling Dissolved Organic Carbon Export from Water Supply Catchments in the Northeastern United States DOI
Kezhen Wang,

Rajith Mukundan,

Rakesh K. Gelda

et al.

Published: Jan. 1, 2024

Natural organic matter (NOM) in rivers is an important energy source to sustain aquatic ecosystem health. However, surface water supply systems where chlorination often used for disinfection, NOM also a precursor the carcinogenic and mutagenic disinfection byproducts such as trihalomethanes haloacetic acids. Effective management of maintain both functions high-quality requires better understanding transport patterns. operationally measured by dissolved carbon (DOC). Challenges using DOC data analysis on catchment scale largely relate spatial temporal variations DOC, low sampling frequency which fails capture multi-scale To help improve sources transport, we analyzed its long-term patterns six catchments New York City Water Supply System monitoring models. We tested empirical models prediction including linear, nonlinear time-series based model formulations. found that generalized additive (GAMs) produced most robust results across catchments. Then, applied calibrated GAM predict daily concentrations estimate fluxes analyze trends. Finally, compared relationships between features investigate regional differences, focusing mechanistic processes associated with parsing out hydrological signals. The showed hydrology plays larger role three top 5% streamflow corresponded nearly 50% annual export, whereas nutrient production were more others. study presents approach streams can inform targeted strategies waters.

Language: Английский

Citations

0