Assessing the Performance of the Gaussian Process Regression Algorithm to Fill Gaps in the Time-Series of Daily Actual Evapotranspiration of Different Crops in Temperate E Continental Zones Using Ground and Remotely Sensed Data DOI
Dario De, Matteo Ippolito,

Marcella Cannarozzo

et al.

Published: Jan. 1, 2023

The knowledge of crop evapotranspiration is crucial for several hydrological processes, including those related to the management agricultural water sources. In particular, estimations actual fluxes within fields are essential managing irrigation strategies save and preserve resources. Among indirect methods estimate evapotranspiration, ETa, eddy covariance (EC) method allows acquire continuous measurement latent heat flux (LE). However, time series EC measurements sometimes characterized by a lack data due sensors' malfunctions. At this aim, Machine Learning (ML) techniques could represent powerful tool fill possible gaps in series. paper, ML technique was applied using Gaussian Process Regression (GPR) algorithm daily evapotranspiration.The tested six different plots, two Italy, three United States America, one Canada, with crops climatic conditions order consider suitability model various contexts. For each site, climate variables were not same, therefore, performance investigated on basis available information. Initially, comparison ground reanalysis data, where both databases available, between satellite products, when have been conducted. Then, GPR tested. mean functions set considering database variables, soil status measurements, remotely sensed vegetation indices. five combinations analyzed verify approach limited input or weather replaced data. Cross-validation used assess procedure. performances assessed based statistical indicators: Root Mean Square Error (RMSE), coefficient determination (R2), Absolute (MAE), regression (b), Nash-Sutcliffe efficiency (NSE). quite high Nash Sutcliffe Efficiency (NSE) coefficient, root square error (RMSE) low values confirm proposed algorithm.

Language: Английский

Enhancing Pan evaporation predictions: Accuracy and uncertainty in hybrid machine learning models DOI Creative Commons
Khabat Khosravi, Aitazaz A. Farooque, Seyed Amir Naghibi

et al.

Ecological Informatics, Journal Year: 2024, Volume and Issue: 85, P. 102933 - 102933

Published: Dec. 7, 2024

Language: Английский

Citations

11
A novel hybrid model combined with ensemble embedded feature selection method for estimating reference evapotranspiration in the North China Plain DOI Creative Commons
Hanmi Zhou,

Linshuang Ma,

Xiaoli Niu

et al.

Agricultural Water Management, Journal Year: 2024, Volume and Issue: 296, P. 108807 - 108807

Published: April 2, 2024

The reference evapotranspiration (ETo) is a key parameter in achieving sustainable use of agricultural water resources. To accurately acquire ETo under limited conditions, this study combined the northern goshawk optimization algorithm (NGO) with extreme gradient boosting (XGBoost) model to propose novel NGO-XGBoost model. performance was evaluated using meteorological data from 30 stations North China Plain and compared XGBoost, random forest (RF), k nearest neighbor (KNN) models. An ensemble embedded feature selection (EEFS) method results RF, adaptive (AdaBoost), categorical (CatBoost) models used obtain importance factors estimating ETo, thereby determine optimal combination inputs indicated that by top 3, 4, 5 important as input combinations, all achieved high estimation accuracy. It worth noting there were significant spatial differences precisions four models, but exhibited consistently precisions, global indicator (GPI) rankings 1st, range coefficient determination (R2), nash efficiency (NSE), root mean square error (RMSE), absolute (MAE) bias (MBE) 0.920–0.998, 0.902–0.998, 0.078–0.623 mm d−1, 0.058–0.430 −0.254–0.062 respectively. Furthermore, accuracy varied across different seasons, which more significantly affected humidity wind speed winter. When target station insufficient, trained historical neighboring still maintained precision. Overall, recommends reliable for provides calculating absence data.

Language: Английский

Citations

10
Machine Learning-Based Estimation of Daily Cropland Evapotranspiration in Diverse Climate Zones DOI Creative Commons

Changmin Du,

Shouzheng Jiang,

Chuqiang Chen

et al.

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(5), P. 730 - 730

Published: Feb. 20, 2024

The accurate prediction of cropland evapotranspiration (ET) is utmost importance for effective irrigation and optimal water resource management. To evaluate the feasibility accuracy ET estimation in various climatic conditions using machine learning models, three-, six-, nine-factor combinations (V3, V6, V9) were examined based on data obtained from global eddy flux sites Moderate Resolution Imaging Spectroradiometer (MODIS) remote sensing data. Four random forest (RF), support vector (SVM), extreme gradient boosting (XGB), backpropagation neural network (BP), used this purpose. input factors included daily mean air temperature (Ta), net radiation (Rn), soil heat (G), evaporative fraction (EF), leaf area index (LAI), photosynthetic photon density (PPFD), vapor pressure deficit (VPD), wind speed (U), atmospheric (P). four models exhibited significant simulation across climate zones, reflected by their performance indicator (GPI) values ranging −3.504 to 0.670 RF, −3.522 1.616 SVM, −3.704 0.972 XGB, −3.654 1.831 BP. choice suitable different varied regions. Specifically, temperate–continental zone (TCCZ), subtropical–Mediterranean (SMCZ), temperate (TCZ), BPC-V9, SVMS-V6, SVMT-V6 demonstrated highest accuracy, with average RMSE 0.259, 0.373, 0.333 mm d−1, MAE 0.177, 0.263, 0.248 R2 0.949, 0.819, 0.917, NSE 0.926, 0.778, 0.899, respectively. In zones a lower LAI there was strong correlation between ET, making more crucial predictions. Conversely, higher (TCZ, SMCZ), significance reduced. This study recognizes impact simulations highlights necessity region-specific considerations when selecting factor combinations.

Language: Английский

Citations

5
Advanced long-term actual evapotranspiration estimation in humid climates for 1958–2021 based on machine learning models enhanced by the RReliefF algorithm DOI Creative Commons
Ahmed Elbeltagi, Salim Heddam, Okan Mert Katipoğlu

et al.

Journal of Hydrology Regional Studies, Journal Year: 2024, Volume and Issue: 56, P. 102043 - 102043

Published: Nov. 1, 2024

Language: Английский

Citations

5
Integrating machine learning and empirical evapotranspiration modeling with DSSAT: Implications for agricultural water management DOI Creative Commons
Niguss Solomon Hailegnaw, Haimanote K. Bayabil, Mulatu Liyew Berihun

et al.

The Science of The Total Environment, Journal Year: 2023, Volume and Issue: 912, P. 169403 - 169403

Published: Dec. 17, 2023

Language: Английский

Citations

12
A practical data-driven approach for precise stem water potential monitoring in pistachio and almond orchards using supervised machine learning algorithms DOI Creative Commons
Mehrad Mortazavi, Stefano Carpin, Arash Toudeshki

et al.

Computers and Electronics in Agriculture, Journal Year: 2025, Volume and Issue: 231, P. 110004 - 110004

Published: Feb. 1, 2025

Language: Английский

Citations

0
State of the Art in Internet of Things Standards and Protocols for Precision Agriculture with an Approach to Semantic Interoperability DOI Creative Commons

Eduard Roccatello,

Antonino Pagano,

Nicolò Levorato

et al.

Network, Journal Year: 2025, Volume and Issue: 5(2), P. 14 - 14

Published: April 21, 2025

The integration of Internet Things (IoT) technology into the agricultural sector enables collection and analysis large amounts data, facilitating greater control over internal processes, resulting in cost reduction improved quality final product. One main challenges designing an IoT system is need for interoperability among devices: different sensors collect information non-homogeneous formats, which are often incompatible with each other. Therefore, user forced to use platforms software consult making complex cumbersome. solution this problem lies adoption standard that standardizes output data. This paper first provides overview standards protocols used precision farming then presents a architecture designed measurements from translate them standard. selected based on state art tailored meet specific needs agriculture. With introduction connector device, can accommodate any number while maintaining data uniform format. Each type sensor associated intercepts intended database translates it format before forwarding central server. Finally, examples real presented illustrate operation connectors their role interoperable architecture, aiming combine flexibility ease low implementation costs.

Language: Английский

Citations

0
An Innovative Approach for Managing the Water Requirements of Fig Trees Using Artificial Intelligence DOI
Josefa Díaz, Francisco P. Chávez, María José Moñino Espino

et al.

Lecture notes in computer science, Journal Year: 2025, Volume and Issue: unknown, P. 273 - 287

Published: Jan. 1, 2025

Language: Английский

Citations

0
Generalization Ability of Bagging and Boosting Type Deep Learning Models in Evapotranspiration Estimation DOI Open Access
Manoranjan Kumar, Yash Agrawal, Sirisha Adamala

et al.

Published: July 5, 2024

The potential of generalized machine learning models developed for crop water estimation was examined in the current study. Extreme Gradient Boosting (XGBoost), Machine (GBM), and Random Forest (RF) are three ensembled that were using all data from a single location 1976 to 2017 then immediately applied at eleven different locations without need any local calibration. For test period January 2018 June 2020, model's capacity estimate numerical values requirement (Pen-man-Monteith (PM) ETo values) assessed. In comparison GBM RF models, XGBoost model outperformed them both marginally significantly. estimate's weighted standard error smaller than 0.85 mm/day, effectiveness varied 96% 99% across various locations. strong performance indicated by decreased noise-to-signal ratio. A real-time management system regional level can be seamlessly linked with this type due its accuracy estimating requirements generalize.

Language: Английский

Citations

3
1D convolutional neural networks-based soil fertility classification and fertilizer prescription DOI

M. Sujatha,

Jaidhar C.D.,

Mallikarjuna Lingappa

et al.

Ecological Informatics, Journal Year: 2023, Volume and Issue: 78, P. 102295 - 102295

Published: Sept. 10, 2023

Language: Английский

Citations

7