Revealing accuracy in climate dynamics: enhancing evapotranspiration estimation using advanced quantile regression and machine learning models

Applied Water Science, Journal Year: 2024, Volume and Issue: 14(7)

Published: June 24, 2024

Abstract This study examines the effectiveness of various quantile regression (QR) and machine learning (ML) methodologies developed for analyzing relationship between meteorological parameters daily reference evapotranspiration (ET ref ) across diverse climates in Iran spanning from 1987 to 2022. The analyzed models include D-vine copula-based (DVQR), multivariate linear (MLQR), Bayesian model averaging (BMAQR), as well algorithms such extreme (ELM), random forest (RF), M5 Tree (M5Tree), least squares support vector algorithm (LSSVR), gradient boosting (XGBoost). Additionally, empirical equations (EEs) Baier Robertson (BARO), Jensen Haise (JEHA), Penman (PENM) were considered. While EEs demonstrated acceptable performance, QR ML exhibited superior accuracy. Among these, MLQR displayed highest accuracy compared DVQR BMAQR models. Moreover, LSSVR, XGBoost, M5Tree outperformed ELM RF Notably, comparable performance (R2 NSE > 0.92, MBE RMSE < 0.5, SI 0.05) all climates. Importantly, these significantly EEs, DVQR, ELM, In conclusion, high-dimensional are recommended promising alternatives accurately estimating ET global climate conditions.

Language: Английский

---

Daily global solar radiation time series prediction using variational mode decomposition combined with multi-functional recurrent fuzzy neural network and quantile regression forests algorithm

Energy Reports, Journal Year: 2023, Volume and Issue: 10, P. 4198 - 4217

Published: Nov. 1, 2023

Global solar radiation (GSR) prediction capability with a reliable model and high accuracy is crucial for comprehending hydrological meteorological systems. It vital the production of renewable clean energy. This research aims to evaluate performance combined variational mode decomposition (VMD) multi-functional recurrent fuzzy neural network (MFRFNN) quantile regression forests (QRF) models GSR in daily scales. The hybrid VMD-MFRFNN QRF were compared standalone MFRFNN, random forest (RF), extreme gradient boosting (XGB), M5 tree (M5T) across Lund Växjö stations Sweden. data from 2008 2017 used train models, while was verified by using 2018 2021 under five different input combinations. various meteorological-based scenarios (including are air temperatures (Tmin, Tmax, T), wind speed (WS), relative humidity (RH), sunshine duration (SSH), maximum possible (N)) considered as predictor models. current study resulted that M5T exhibited higher than RF XGB showed equivalent at both sites. MFRFNN outperformed all combinations best when fewer variables T, WS station Tmin, WS, SSH, RH station) prediction. We conclude predicts average combining RH, N).

Language: Английский

---

High performance machine learning approach for reference evapotranspiration estimation

Stochastic Environmental Research and Risk Assessment, Journal Year: 2023, Volume and Issue: 38(2), P. 689 - 713

Published: Nov. 4, 2023

Abstract Accurate reference evapotranspiration (ET 0 ) estimation has an effective role in reducing water losses and raising the efficiency of irrigation management. The complicated nature process is illustrated amount meteorological variables required to estimate ET . Incomplete data most significant challenge that confronts estimation. For this reason, different machine learning techniques have been employed predict , but structures architectures many them make very difficult. these challenges, ensemble are frequently for estimating particularly when there a shortage data. This paper introduces powerful super learner technique estimation, where four models: Extra Tree Regressor, Support Vector K-Nearest Neighbor AdaBoost Regression represent base learners their outcomes used as training meta learner. Overcoming overfitting problem affects other methods advantage cross-validation theory-based approach. Super performances were compared with forecasting capabilities through statistical standards, results revealed better accuracy than learners, combinations whereas Coefficient Determination (R 2 ranged from 0.9279 0.9994 Mean Squared Error (MSE) 0.0026 0.3289 mm/day R 0.5592 0.9977, MSE 0.0896 2.0118 therefore, highly recommended prediction limited

Language: Английский

---

Estimating reference crop evapotranspiration using optimized empirical methods with a novel improved Grey Wolf Algorithm in four climatic regions of China

Agricultural Water Management, Journal Year: 2023, Volume and Issue: 291, P. 108620 - 108620

Published: Dec. 12, 2023

Accurate estimation of reference crop evapotranspiration (ETo) is crucial for agricultural water management. As the simplified alternatives Penman-Monteith equation, empirical methods have been widely recommended worldwide. However, its application still limited to parameters localization varied with geographical and climatic conditions, therefore developing an excellent optimization algorithm calibrating very necessary. Regarding above requirement, present study developed a novel improved Grey Wolf Algorithm (MDSL-GWA) optimize most ones among three types ETo methods. After performance comparison Least Square Method (LSM), Genetic (GA), (GWA), MDSL-GWA in four regions China, this found that Priestley-Taylor (PT) method was best radiation-based (Rn-based) achieved better temperate continental region (TCR), mountain plateau (MPR), monsoon (TMR) than other types. While temperature-based (T-based) Hargreaves-Samani (HS) performed subtropical (SMR), further attaching same type TMR TCR, while Oudin T-based MPR. Moreover, Romanenko humidity-based (RH-based) TCR MPR, whereas Brockamp-Wenner exhibited higher SMR TMR. Furthermore, despite intelligence algorithms significantly enhancing original methods, outperformed by 4.5–29.6% determination coefficient (R2), 4.7–27.3% nash-sutcliffe efficient (NSE), 3.7–44.4% relative root mean square error (RRMSE), 3.1–56.2% absolute (MAE), respectively. optimization, MDSL-GWA-PT TMR, median values R2, NSE, RRMSE, MAE ranged 0.907–0.958, 0.887–0.925, 0.083–0.103, 0.115–0.162 mm, In SMR, MDSL-GWA-HS produced estimates, being 0.876, 0.843, 0.112, 0.146 summary, using accessible data which helpful decision-making effective management utilization regional resources.

Language: Английский

---

Wear Prediction of Functionally Graded Composites Using Machine Learning

Materials, Journal Year: 2024, Volume and Issue: 17(18), P. 4523 - 4523

Published: Sept. 14, 2024

This study focuses on the production of functionally graded composites by utilizing magnesium matrix waste chips and cost-effective eggshell reinforcements through centrifugal casting. The wear behavior produced samples was thoroughly examined, considering a range loads (5 N to 35 N), sliding speeds (0.5 m/s 3.5 m/s), distances (500 m 3500 m). worn surfaces were carefully analyzed gain insights into underlying mechanisms. results indicated successful particle integration in levels within composite, enhancing hardness resistance. In outer zone, there 25.26% increase over inner zone due gradient, with resistance improving 19.8% compared zone. To predict behavior, four distinct machine learning algorithms employed, their performance using limited dataset obtained from various test operations. tree-based model surpassed deep neural-based models predicting rate among developed models. These provide fast effective way evaluate reinforced particles for specific applications, potentially decreasing need extensive additional tests. Notably, LightGBM exhibited highest accuracy testing set across three zones. Finally, findings highlighted viability employing crafting composites. approach not only minimizes environmental impact material repurposing but also offers means these resources creating automotive components that demand varying properties surfaces, regions.

Language: Английский

---

An intelligent and uncertain optimization framework for water-nitrogen synergistic management under extreme supply and demand water risks

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 127829 - 127829

Published: April 1, 2025

Language: Английский

---

Rapid and non-destructive monitoring of the drying process of glutinous rice using visible-near infrared hyperspectral imaging

Applied Food Research, Journal Year: 2025, Volume and Issue: unknown, P. 100955 - 100955

Published: May 1, 2025

Language: Английский

---

Beyond the turbine: Charting the ecological footprint trajectory of wind energy technology budgets

Energy & Environment, Journal Year: 2025, Volume and Issue: unknown

Published: May 16, 2025

Amid the world's pursuit of environmental responsibility, strategic investments in wind energy technology reveal a powerful synergy, illuminating path toward greener and more sustainable future. This research explores asymmetric association between budgets ecological footprint ten leading nations that invest most R&D ( USA, China, Italy, UK, Brazil, France, India, Spain, Canada, Germany ). Prior investigations utilized panel data approaches to probe technology-environment nexus without accounting for specific qualities various economies. Contrarily, current adopts Quantile-on-Quantile methodology appraise this relationship individually every nation. unique approach improves exactness our estimation, delivering holistic global viewpoint while customized perceptions particular economy. The annual economies extends 2000 2023. findings indicate dedicating resources quality by reducing across several quantiles selected counties. Furthermore, highlight diverse behaviors these linkages sample These results underline significance policymakers performing exhaustive appraisals executing efficient measures allocate sustainability. Highlights study analyzes technology-ecological nexus. A methodology, “Quantile-on-Quantile (QQ),” is used. Wind reduces footprint.

Language: Английский

---

Investigating the probabilistic behavior of reference evapotranspiration using Vine tree sequence

Applied Water Science, Journal Year: 2025, Volume and Issue: 15(5)

Published: April 18, 2025

Abstract Reference evapotranspiration, which includes the contribution of climatic conditions in potential is considered as an important and strategic criterion water resources management irrigation designs. Therefore, it necessary to determine predict its changes each region. In this study, using copula functions, behavior component were investigated west Iran. For purpose, meteorological information nine synoptic stations including Tmax, Tmin, WS, Rs, RHmax, RHmin used. This research aims explore multivariate simulation based on vine tree sequences. Among these parameters, wind speed had least effect ET 0 , all studied stations, there was highest correlation between -Tmax pair variable, equal 0.90, 0.87, 0.89, 0.88, 0.86, 0.85, 0.81 Aligudarz, Azna, Borujerd, Dorud, Khorramabad, Kuhdasht, Nurabad, Poldakhter respectively, Kendall's Tau statistics. The sequence copulas C-, D-, R-vine examined according input variables AIC logarithm likelihood evaluation criteria. According results, found that criteria, D-vine has best performance joint probability analysis variables. addition, results showed sequence, unlike two R C type sequences, maintained until last tree. study functions could analyze evapotranspiration different climates with high capability, can be used predicting non-linear

Language: Английский

---

A D-vine copula-based quantile regression towards merging satellite precipitation products over rugged topography: a case study in the upper Tekeze–Atbara Basin

Hydrology and earth system sciences, Journal Year: 2024, Volume and Issue: 28(5), P. 1147 - 1172

Published: March 7, 2024

Abstract. Precipitation is a vital key element in various studies of hydrology, flood prediction, drought monitoring, and water resource management. The main challenge conducting over remote regions with rugged topography that weather stations are usually scarce unevenly distributed. However, open-source satellite-based precipitation products (SPPs) suitable resolution provide alternative options these data-scarce regions, which typically associated high uncertainty. To reduce the uncertainty individual satellite products, we have proposed D-vine copula-based quantile regression (DVQR) model to merge multiple SPPs rain gauges (RGs). DVQR was employed during 2001–2017 summer monsoon seasons compared two other methods based on multivariate linear (MLQR) Bayesian averaging (BMAQ) techniques, respectively, traditional merging – simple modeling average (SMA) one-outlier-removed (OORA) using descriptive categorical statistics. Four been considered this study, namely, Tropical Applications Meteorology SATellite (TAMSAT v3.1), Climate Prediction Center MORPHing Product Data Record (CMORPH-CDR), Global Measurement (GPM) Integrated Multi-satellitE Retrievals for GPM (IMERG v06), Estimation from Remotely Sensed Information Artificial Neural Networks (PERSIANN-CDR). bilinear (BIL) interpolation technique applied downscale coarse fine spatial (1 km). rugged-topography region upper Tekeze–Atbara Basin (UTAB) Ethiopia selected as study area. results indicate data estimates DVQR, MLQR, BMAQ models outperform downscaled SPPs. Monthly evaluations reveal all perform better July September than June August due variability. exhibit higher accuracy UTAB. substantially improved statistical metrics (CC = 0.80, NSE 0.615, KGE 0.785, MAE 1.97 mm d−1, RMSE 2.86 PBIAS 0.96 %) MLQR models. did not respect probability detection (POD) false-alarm ratio (FAR), although it had best frequency bias index (FBI) critical success (CSI) among Overall, newly approach improves quality demonstrates value such

Language: Английский

---