Improving potential evapotranspiration prediction accuracy using advanced machine learning models: a case study in Ahvaz, Iran

Jamshid Bani Feri; Aslan Egdernezhad; Ali Mokhtaran; Mahdi Asadilour; Davoud Khodadadi Dehkordi; Jamshid Bani Feri; Aslan Egdernezhad; Ali Mokhtaran; Mahdi Asadilour; Davoud Khodadadi Dehkordi

doi:10.3934/environsci.2025034

AIMS Environmental Science

2025, Volume 12, Issue 5: 770-794. doi: 10.3934/environsci.2025034

Previous Article Next Article

Research article Special Issues

Improving potential evapotranspiration prediction accuracy using advanced machine learning models: a case study in Ahvaz, Iran

1.
Ph.D. student, Department of Water Sciences and Engineering, Ahv.C., Islamic Azad University, Ahvaz, Iran; jamshidbaniferi@iau.ac.ir
2.
Assistant Professor, Department of Water Sciences and Engineering, Ahv.C., Islamic Azad University, Ahvaz, Iran; a_eigder@iau.ac.ir
3.
Assistant Professor., Agricultural Engineering Research Department, Khuzestan Agricultural and Natural Resources Research and Education Center, Agricultural Research Education and Extension Organization (AREEO), Ahwaz, Iran; alimokhtaran@gmail.com
4.
Assistant Professor, Department of Water Sciences and Engineering, Ahv.C., Islamic Azad University, Ahvaz, Iran; mehdi.asadilour@iau.ac.ir
5.
Assistant Professor, Department of Water Sciences and Engineering, Ahv.C., Islamic Azad University, Ahvaz, Iran; korkori@iau.ac.ir

Received: 11 December 2024 Revised: 03 July 2025 Accepted: 28 July 2025 Published: 14 August 2025

Accurate forecasting of daily potential evapotranspiration (ET_o) is vital for efficient irrigation scheduling, sustainable water resource management, and optimal crop yield, especially in arid regions. In this study, we present a novel, large-scale comparison of three advanced machine learning models, Support Vector Machine (SVM), Artificial Neural Network (ANN), and Gene Expression Programming (GEP), for ET_o prediction at the Ahvaz synoptic station in Iran. The work is distinctive in using one of the longest ET_o datasets available (1979–2023; 16,084 records), incorporating a sensitivity analysis for input selection, and applying the Developed Discrepancy Ratio (DDR) as an advanced performance metric. The ANN model (MLP 4-9-1 architecture) demonstrated the highest prediction accuracy, achieving an R² of 0.9806, RMSE of 0.4122, and DDRmax of 3.27 in the training phase, and an R² of 0.9779, RMSE of 0.4327, and DDR_max of 3.22 during validation. In comparison, SVM and GEP models showed lower accuracy across all phases. These results highlight the superior capability of the ANN model for ET_o forecasting and its potential as a reliable tool for irrigation planning and water resource management in arid regions like Ahvaz.
- climate,
- artificial intelligence,
- dry region,
- water cycle,
- performance assessment
Citation: Jamshid Bani Feri, Aslan Egdernezhad, Ali Mokhtaran, Mahdi Asadilour, Davoud Khodadadi Dehkordi. Improving potential evapotranspiration prediction accuracy using advanced machine learning models: a case study in Ahvaz, Iran[J]. AIMS Environmental Science, 2025, 12(5): 770-794. doi: 10.3934/environsci.2025034

Related Papers:

Abstract

Accurate forecasting of daily potential evapotranspiration (ET_o) is vital for efficient irrigation scheduling, sustainable water resource management, and optimal crop yield, especially in arid regions. In this study, we present a novel, large-scale comparison of three advanced machine learning models, Support Vector Machine (SVM), Artificial Neural Network (ANN), and Gene Expression Programming (GEP), for ET_o prediction at the Ahvaz synoptic station in Iran. The work is distinctive in using one of the longest ET_o datasets available (1979–2023; 16,084 records), incorporating a sensitivity analysis for input selection, and applying the Developed Discrepancy Ratio (DDR) as an advanced performance metric. The ANN model (MLP 4-9-1 architecture) demonstrated the highest prediction accuracy, achieving an R² of 0.9806, RMSE of 0.4122, and DDRmax of 3.27 in the training phase, and an R² of 0.9779, RMSE of 0.4327, and DDR_max of 3.22 during validation. In comparison, SVM and GEP models showed lower accuracy across all phases. These results highlight the superior capability of the ANN model for ET_o forecasting and its potential as a reliable tool for irrigation planning and water resource management in arid regions like Ahvaz.

References

[1]	Braddock RD, Kremmer ML, Sanzogni, L (1998) Feed‐forward artificial neural network model for forecasting rainfall run‐off. Environmetrics 9: 419–432. https://doi.org/10.1002/(SICI)1099-095X(199807/08)9:4<419::AID-ENV312>3.0.CO;2-D doi: 10.1002/(SICI)1099-095X(199807/08)9:4<419::AID-ENV312>3.0.CO;2-D
[2]	Tripathi S, Srinivas VV, Nanjundiah RS (2006) Downscaling of precipitation for climatechange scenarios: a support vector machine approach. J hydrol 330: 621–640. https://doi.org/10.1016/j.jhydrol.2006.04.030 doi: 10.1016/j.jhydrol.2006.04.030
[3]	Nash JE, Sutcliffe JV (1970) River flow forecasting through conceptual models part Ⅰ—A discussion of principles. J hydrol 10: 282–290. https://doi.org/10.1016/0022-1694(70)90255-6 doi: 10.1016/0022-1694(70)90255-6
[4]	Zhang G, Bateni SM, Jun C, Khoshkam H, Band SS, Mosavi A (2022) Feasibility of random forest and multivariate adaptive regression splines for predicting long-term mean monthly dew point temperature. Front Environ Sci 10: 826165. https://doi.org/10.3389/fenvs.2022.826165 doi: 10.3389/fenvs.2022.826165
[5]	Elbeltagi A, Kushwaha NL, Rajput J, et al. (2022) Modelling daily reference evapotranspiration based on stacking hybridization of ANN with meta-heuristic algorithms under diverse agro-climatic conditions. Stoch Environ Res Risk Assess 36: 3311–3334. https://doi.org/10.1007/s00477-022-02196-0 doi: 10.1007/s00477-022-02196-0
[6]	Dimitriadou S, Nikolakopoulos KG (2022) Artificial neural networks for the prediction of the reference evapotranspiration of the Peloponnese Peninsula, Greece. Water 14: 2027. https://doi.org/10.3390/w14132027 doi: 10.3390/w14132027
[7]	Malik A, Jamei M, Ali M, et al. (2022) Multi-step daily forecasting of reference evapotranspiration for different climates of India: A modern multivariate complementary technique reinforced with ridge regression feature selection. Agr Water Manage 272: 107812. https://doi.org/10.1016/j.agwat.2022.107812 doi: 10.1016/j.agwat.2022.107812
[8]	Aghelpour P, Norooz-Valashedi R (2022) Predicting daily reference evapotranspiration rates in a humid region, comparison of seven various data-based predictor models. Stoch Environ Res Risk A 36: 4133–4155. https://doi.org/10.1007/s00477-022-02249-4 doi: 10.1007/s00477-022-02249-4
[9]	Achite M, Jehanzaib M, Sattari MT, et al. (2022) Modern techniques to modeling reference evapotranspiration in a semiarid area based on ANN and GEP models. Water 14: 1210. https://doi.org/10.3390/w14081210 doi: 10.3390/w14081210
[10]	Sanikhani H, Deo RC, Samui P, et al. (2018) Survey of different data-intelligent modeling strategies for forecasting air temperature using geographic information as model predictors. Comput Electron Agr 152: 242–260. https://doi.org/10.1016/j.compag.2018.07.008 doi: 10.1016/j.compag.2018.07.008
[11]	Bellido-Jiménez JA, Gualda JE, García-Marín AP, (2021) Assessing new intra-daily temperature-based machine learning models to outperform solar radiation predictions in different conditions. Appl Energ 298: 117211. https://doi.org/10.1016/j.apenergy.2021.117211 doi: 10.1016/j.apenergy.2021.117211
[12]	Yakut E, Süzülmüş S (2020) Modelling monthly mean air temperature using artificial neural network, adaptive neuro-fuzzy inference system and support vector regression methods: A case of study for Turkey. Network Comp Neural 31: 1–36. https://doi.org/10.1080/0954898X.2020.1759833 doi: 10.1080/0954898X.2020.1759833
[13]	Citakoglu H, Babayigit B, Haktanir NA, (2020) Solar radiation prediction using multi-gene genetic programming approach. Theor Appl Climatol 142: 885–897. DOI: 10.1007/s00704-020-03356-4 doi: 10.1007/s00704-020-03356-4
[14]	Biazar SM, Rahmani V, Isazadeh M, et al. (2020) New input selection procedure for machine learning methods in estimating daily global solar radiation. Arab J Geosci 13: 1–17. https://doi.org/10.1007/s12517-020-05437-0 doi: 10.1007/s12517-020-05437-0
[15]	Raza A, Shoaib M, Faiz MA, et al. (2020). Comparative assessment of reference evapotranspiration estimation using conventional method and machine learning algorithms in four climatic regions. Pure Appl Geophys 177: 4479–4508. https://doi.org/10.1007/s00024-020-02473-5 doi: 10.1007/s00024-020-02473-5
[16]	Lu X, Fan J, Wu L, et al. (2020) Forecasting multi-step ahead monthly reference evapotranspiration using hybrid extreme gradient boosting with grey wolf optimization algorithm. Comp Model Eng Sci 125: 699–723. https://doi.org/10.32604/cmes.2020.011004 doi: 10.32604/cmes.2020.011004
[17]	Maroufpoor S, Sanikhani H, Kisi O, et al. (2019) Long‐term modelling of wind speeds using six different heuristic artificial intelligence approaches. Int J Climatol 39: 3543–3557. https://doi.org/10.1002/joc.6037 doi: 10.1002/joc.6037
[18]	Jahani B, Mohammadi B, (2019) A comparison between the application of empirical and ANN methods for estimation of daily global solar radiation in Iran. Theor Appl Climatol 137: 1257–1269. https://doi.org/10.1007/s00704-018-2666-3 doi: 10.1007/s00704-018-2666-3
[19]	Kiafar H, Babazadeh H, Marti P, et al. (2017) Evaluating the generalizability of GEP models for estimating reference evapotranspiration in distant humid and arid locations. Theor Appl Climatol 130: 377–389. https://doi.org/10.1007/s00704-016-1888-5 doi: 10.1007/s00704-016-1888-5
[20]	Mehdizadeh S, Behmanesh J, Khalili K (2017) Using MARS, SVM, GEP and empirical equations for estimation of monthly mean reference evapotranspiration. Comput Electron Agr 139: 103–114. https://doi.org/10.1016/j.compag.2017.05.002 doi: 10.1016/j.compag.2017.05.002
[21]	Tezel G, Buyukyildiz M (2016). Monthly evaporation forecasting using artificial neural networks and support vector machines. Theor Appl Climatol 124: 69–80. https://doi.org/10.1007/s00704-015-1392-3 doi: 10.1007/s00704-015-1392-3
[22]	Sharifi SS, Rezaverdinejad V, Nourani V (2016) Estimation of daily global solar radiation using wavelet regression, ANN, GEP and empirical models: A comparative study of selected temperature-based approaches. J Atmos Sol-Terr Phys 149: 131–145. https://doi.org/10.1016/j.jastp.2016.10.008 doi: 10.1016/j.jastp.2016.10.008
[23]	Martí P, González-Altozano P, López-Urrea R, et al. (2015) Modeling reference evapotranspiration with calculated targets. Assessment and implications. Agr Water Manage 149: 81–90. https://doi.org/10.1016/j.agwat.2014.10.028 doi: 10.1016/j.agwat.2014.10.028
[24]	Kisi O, Sanikhani H, Zounemat-Kermani M, et al. (2015) Long-term monthly evapotranspiration modeling by several data-driven methods without climatic data. Comput Electron Agr 115: 66–77. https://doi.org/10.1016/j.compag.2015.04.015 doi: 10.1016/j.compag.2015.04.015
[25]	Shiri J, Marti P, Kisi O, et al. (2013) Evaluation of different data management scenarios for estimating daily reference evapotranspiration. Hydrol Res 44: 1058–1070. https://doi.org/10.2166/nh.2013.154 doi: 10.2166/nh.2013.154
[26]	Traore S, Guven A, (2012) Regional-specific numerical models of evapotranspiration using gene-expression programming interface in sahel. Water Resour Manag 26: 4367–4380. https://doi.org/10.1007/s11269-012-0149-3 doi: 10.1007/s11269-012-0149-3
[27]	Izadifar Z, Elshorbagy A, (2010) Prediction of hourly actual evapotranspiration using neural networks, genetic programming, and statistical models. Hydrol Process 24: 3413–3425. https://doi.org/10.1002/hyp.7771 doi: 10.1002/hyp.7771
[28]	Kisi O, Guven A, (2010) Evapotranspiration modeling using linear genetic programming technique. J Irrig Drain Eng 136: 715–723. https://doi.org/10.1061/(ASCE)IR.1943-4774.0000244 doi: 10.1061/(ASCE)IR.1943-4774.0000244
[29]	Kim S, Kim HS, (2008) Neural networks and genetic algorithm approach for nonlinear evaporation and evapotranspiration modeling. J Hydrol 351. https://doi.org/10.1016/j.jhydrol.2007.12.014 doi: 10.1016/j.jhydrol.2007.12.014
[30]	Fuladipanah M, Azamathulla HM, Kisi O, et al. (2024) Quantitative forecasting of bed sediment load in river engineering: an investigation into machine learning methodologies for complex phenomena. Water Supply 24: 585-600. https://doi.org/10.2166/ws.2024.017 doi: 10.2166/ws.2024.017
[31]	Sajindra H, Abekoon T, Wimalasiri EM, et al. (2023) An Artificial Neural Network for Predicting Groundnut Yield Using Climatic Data. Agri Engineering 5 1713-1736. https://doi.org/10.3390/agriengineering5040106 doi: 10.3390/agriengineering5040106
[32]	Jayathilake T, Sarukkalige R, Hoshino Y, et al. (2022) Wetland water level prediction using artificial neural networks-A case study in the Colombo flood detention area, Sri Lanka. Climate 11. https://doi.org/10.3390/cli11010001 doi: 10.3390/cli11010001
[33]	Azamathulla HM, Rathnayake U, Shatnawi A (2018) Gene expression programming and artificial neural network to estimate atmospheric temperature in Tabuk, Saudi Arabia. Appl Water Sci 8: 1-7. https://doi.org/10.1007/s13201-018-0831-6 doi: 10.1007/s13201-018-0831-6
[34]	Leon LP, Azamathulla HM, Felix P, et al. (2023). Prediction of stiffness modulus of bituminous mixtures using the applications of multi-expression programming and gene expression programming. Road Mater Pavement Des 24: 2192-2208. https://doi.org/10.1080/14680629.2022.2126383 doi: 10.1080/14680629.2022.2126383
[35]	Fuladipanah M, Hazi MA, Kisi O, (2023) An in-depth comparative analysis of data-driven and classic regression models for scour depth prediction around cylindrical bridge piers. Appl Water Sci 13: 231. https://doi.org/10.1007/s13201-023-02022-0 doi: 10.1007/s13201-023-02022-0
[36]	Gharehbaghi A, Ghasemlounia R, Afaridegan E, et al. (2023) A comparison of artificial intelligence approaches in predicting discharge coefficient of streamlined weirs. J Hydroinform 25: 1513-1530. https://doi.org/10.2166/hydro.2023.063 doi: 10.2166/hydro.2023.063
[37]	Azamathulla HM (2012). Gene-expression programming to predict scour at a bridge abutment. J Hydroinform 14: 324-331. https://doi.org/10.2166/hydro.2011.135 doi: 10.2166/hydro.2011.135
[38]	Cortes C., Vapnik V. (1995). Support-vector networks. Mach learn 20: 273-297. https://doi.org/10.1007/BF00994018 doi: 10.1007/BF00994018
[39]	Kumar V, Azamathulla HM, Sharma KV, et al. (2023) The state of the art in deep learning applications, challenges, and future prospects: A comprehensive review of flood forecasting and management. Sustainability 15: 10543. https://doi.org/10.3390/su151310543 doi: 10.3390/su151310543
[40]	Rathnayake N, Rathnayake U, Dang TL, et al. (2023) Water level prediction using soft computing techniques: A case study in the Malwathu Oya, Sri Lanka. Plos one 18: e0282847. https://doi.org/10.1371/journal.pone.0282847 doi: 10.1371/journal.pone.0282847
[41]	Noori R, Khakpour A, Omidvar B, et al. (2010) Comparison of ANN and principal component analysis-multivariate linear regression models for predicting the river flow based on developed discrepancy ratio statistics. Expert Syst Appl 37: 5856-5862. https://doi.org/10.1016/j.eswa.2010.02.020 doi: 10.1016/j.eswa.2010.02.020
[42]	Koncar N, (1997). Optimization methodologies for direct inverse neurocontrol. (Publication No., SW72BZ) [Doctoral dissertation, London University, Englan].
[43]	Malik A, Tikhamarine Y, Al-Ansari N, et al. (2021) Daily pan-evaporation estimation in different agroclimatic zones using novel hybrid support vector regression optimized by Salp swarm algorithm in conjunction with gamma test. Eng Appl Comp Fluid Mech 15: 1075-1094. https://doi.org/10.1080/19942060.2021.1942990 doi: 10.1080/19942060.2021.1942990

Reader Comments

Your name:*

Email:*
© 2025 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)