Predicting spatial distribution of stable isotopes in precipitation by classical geostatistical- and machine learning methods

Journal of Hydrology, Journal Year: 2023, Volume and Issue: 617, P. 129129 - 129129

Published: Jan. 16, 2023

Stable isotopes of precipitation are important natural tracers in hydrology, ecology, and forensics. The spatially explicit predictions oxygen hydrogen obtained through different interpolation techniques. In the present study we aim to examine performance various techniques when predicting spatial distribution stable isotopes. efficiency combined geostatistical tools (i.e. regression kriging; RK) machine learning methods (including enhanced random forest methods: MRRF, RERF) compared interpolating variability isotope values from two sampling networks Europe. To assess models, mean squared error (MSE), nonparametric Kling Gupta (KGE), absolute differences relative metrics were employed. It was found that combination with Random Forest can produce estimations comparable accuracy terms descending order overall average MSE, MRRF: 2.61, RK: 2.77, RERF: 2.99, RF: 3.08. best performing model variant (MRRF) outperformed kriging a hybrid metric (KGE) by 7.5%. Sequential rarefying station showed machine-learning more capable maintaining high prediction even fewer input data. This be great advantage suitable method is needed predict composition for large domains where density data stations shows differences.

Language: Английский

---

A 3-D groundwater isoscape of the contiguous USA for forensic and water resource science

PLoS ONE, Journal Year: 2022, Volume and Issue: 17(1), P. e0261651 - e0261651

Published: Jan. 7, 2022

A wide range of hydrological, ecological, environmental, and forensic science applications rely on predictive “isoscape” maps to provide estimates the hydrogen or oxygen isotopic compositions environmental water sources. Many isoscapes have been developed, but few studies produced specifically representing groundwaters. None these represented distinct subsurface layers variations across them. Here we compiled >6 million well completion records >27,000 groundwater isotope datapoints develop a space- depth-explicit isoscape for contiguous United States. This 3-dimensional model shows that vertical heterogeneity in is substantial some parts country delta values often differ from those coincident precipitation surface resources; many patterns can be explained by established hydrological hydrogeological mechanisms. We validate against an independent data set tap show accurately predicts communities known use resources. new approach represents foundation further developments resulting should improved predictions systems where potential source.

Language: Английский

---

Linking precipitation and groundwater isotopes in Ethiopia - Implications from local meteoric water lines and isoscapes

Journal of Hydrology, Journal Year: 2021, Volume and Issue: 596, P. 126074 - 126074

Published: Feb. 16, 2021

Language: Английский

---

Shift in groundwater recharge of the Bengal Basin from rainfall to surface water

Communications Earth & Environment, Journal Year: 2023, Volume and Issue: 4(1)

Published: Jan. 9, 2023

Abstract Groundwater supports agriculture and provides domestic water for over 250 million people in the Bengal Basin. Here we investigate source of groundwater recharge using 2500 stable isotope measurements from region. We employ a Monte Carlo statistical analysis to find distributions possible components by accounting variability ratios each sources. that sources have shifted last decades with ~50% increase stagnant surface bodies (mostly during latter part dry season) relative decrease contribution direct infiltration precipitation (which occurs mostly early monsoon). attribute this shift an standing irrigated rice fields ponds, downward hydraulic gradient season driven pumping.

Language: Английский

---

Quantitative expression of LNAPL pollutant concentrations in capillary zone by coupling multiple environmental factors based on random forest algorithm

Journal of Hazardous Materials, Journal Year: 2024, Volume and Issue: 479, P. 135695 - 135695

Published: Aug. 28, 2024

Language: Английский

---

Investigating the connection between the global meteoric water line and the local meteoric water line within the Gauteng and North West provinces in South Africa

Journal of Hydrology Regional Studies, Journal Year: 2025, Volume and Issue: 58, P. 102302 - 102302

Published: March 10, 2025

Language: Английский

---

Predicting the spatial distribution of stable isotopes in precipitation using a machine learning approach: a comparative assessment of random forest variants

GEM - International Journal on Geomathematics, Journal Year: 2023, Volume and Issue: 14(1)

Published: June 12, 2023

Abstract Stable isotopes of hydrogen and oxygen are important natural tracers with a wide variety environmental applications (e.g., the exploration water cycle, ecology food authenticity). The spatially explicit predictions their variations obtained through various interpolation techniques. In present work, classical random forest (RF) two its variants were applied. RF version employing buffer distance (RF sp ) applied to each month separately, while model was trained using all data year as categorical variables tg ). Their performance in predicting spatial variability precipitation stable isotope values for 2008–2017 across Europe compared. addition, comparison made publicly available alternative machine learning which employs extreme gradient boosting. Input retrieved from Global Network Isotopes Precipitation (GNIP; no. stations: 144) other national datasets (no. 127). Comparisons on basis absolute differences, median, mean error Lin’s concordance correlation coefficient. All capable reproducing overall trends seasonal patterns over time measured at chosen validation site Europe. most predictors latitude case RF, meteorological (vapor pressure, saturation vapor temperature) models. Diurnal temperature range had weakest predictive power every case. conclusion, it may be stated that merged dataset, combining GNIP datasets, yielded smallest 1.345‰) highest coefficient (0.987), boosting (based only data) 1.354‰, 0.984, although produced lowers median value (1.113‰), 1.124‰. striking systematic bias observed summer season northern stations; this, however, diminished 2014 onward, point after stations beyond 55° N training set.

Language: Английский

---

Climate Impacts on Source Contributions and Evaporation to Flow in the Snake River Basin Using Surface Water Isoscapes (δ²H and δ¹⁸O)

Water Resources Research, Journal Year: 2021, Volume and Issue: 57(7)

Published: June 21, 2021

Rising global temperatures are expected to decrease the precipitation amount that falls as snow, causing greater risk of water scarcity, groundwater overdraft, and fire in areas rely on mountain snowpack for their supply. Streamflow large river basins varies with amount, timing, type precipitation, evapotranspiration, drainage properties watersheds; however, these controls vary time space making it difficult identify contributing most flow when. In this study, we separate evaporative influences from source values isotopes Snake River Basin western United States (US) relate area dynamics. We developed isoscapes (δ2H δ18O) basin found isotopic composition surface small watersheds is primarily controlled by longitude, latitude, elevation. To examine temporal variability contributions flow, present a six-year record King Hill, Idaho after removing influences. During periods low were isotopically lighter indicating larger contribution waters highest elevation, eastern portion basin. evaporation increases evident during summer likely reflecting climate, changing availability, management strategies within Our findings potential tool identifying critical portions climate fluctuations alter This can be applied other continental-interior where may obscure signatures.

Language: Английский

---

Seasonal recharge mechanism of the upper shallow groundwater in a long-term wastewater leakage and irrigation region of an alluvial aquifer

Journal of Hydrology, Journal Year: 2023, Volume and Issue: 629, P. 130424 - 130424

Published: Nov. 6, 2023

Language: Английский

---

Using Stable Isotopes to Determine the Water Balance of Utah Lake (Utah, USA)

Hydrology, Journal Year: 2020, Volume and Issue: 7(4), P. 88 - 88

Published: Nov. 16, 2020

To investigate the hydrology of Utah Lake, we analyzed hydrogen (δ2H) and oxygen (δ18O) stable isotope composition water samples collected from various components its system. The average δ2H δ18O values inlets are similar to groundwater, which in turn has a that is winter precipitation. This suggests snowmelt-fed groundwater main source Valley river waters. In addition, plot close local meteoric line, suggesting no significant evaporation occurring these rivers. contrast, lake outlet have higher δ-values than along lines, occurrence evaporation. Isotope data also indicate poorly mixed horizontally, but well vertically. Calculations based on mass balance equations provide estimates for percentage input lost by (~47%), residence time (~0.5 years), volume inflow (~700 million m3) during period April November. short high total coming might suggest more susceptible pollution surface pollution.

Language: Английский

---