Correo Electrónico
DNINFOA - SIA
Bibliotecas
Convocatorias
Identidad U.N.
Published
Machine Learning Models for Evaluation and Optimizing Evapotranspiration Prediction
Modelos de aprendizaje automático para evaluar y optimizar la predicción de la evapotranspiración
DOI:
https://doi.org/10.15446/ing.investig.113275Keywords:
evapotranspiration, machine learning, ANFIS, GEP, M5P tree (en)evapotranspiración, aprendizaje automático, ANFIS, GEP, árbol M5P (es)
Downloads
This study extensively analyzed three models, i.e., M5P (model tree), ANFIS (adaptive neuro-fuzzy inference system), and GEP (gene expression programming), to predict actual evapotranspiration (ETo) at six major stations in the Mahanadi Basin region: Raipur, Korba, Jharsuguda, Bilaspur, Bhubaneswar, and Balangir. Evaluation metrics including the R2, RMSE, NSE, and MAE were applied to the testing dataset, revealing ANFIS's consistent superiority, with high R2 (0.930746-0,990526) and NSE (0.926792-0.990458) values alongside the lowest RMSE (0.101152-0.332819) and MAE (0.000386-0.034319). Weighted scores confirmed ANFIS's dominance across multiple stations, except for specific instances: GEP excelled in Bhubaneswar and M5P in Balangir. This study highlighted ANFIS's proficiency in predicting ETo values across various locations, as demonstrated through the effective capture of variations in scatterplots. This underscores the importance of model selection, the versatility of machine learning models, and the effectiveness of combining artificial intelligence techniques for accurate ETo prediction. ANFIS consistently outperformed M5P and GEP, solidifying its status as a reliable prediction tool. While acknowledging the potential of M5P and GEP in specific contexts, this study stresses the need to tailor models to unique location characteristics. References to related studies supported the effectiveness of hybridized AI approaches in improving ETo modeling. We encourage ongoing research to refine the models, incorporate additional factors, and enhance predictive accuracy. Our findings provide valuable insights for water resource management, irrigation planning, and agricultural decision-making across diverse locations.
Este estudio analizó de manera exhaustiva tres modelos, i.e., el árbol de modelo (M5P), el sistema de inferencia neuro-difuso adaptativo (ANFIS) y la programación de expresión génica (GEP), para predecir la evapotranspiración real (ETo) en seis estaciones principales de la cuenca del Mahanadi: Raipur, Korba, Jharsuguda, Bilaspur, Bhubaneswar y Balangir. Se aplicaron métricas de evaluación al conjunto de prueba, incluyendo el R², el RMSE, la NSE y el MAE, lo que reveló la consistente superioridad de ANFIS, con valores altos de R² (0.930746-0.990526) y NSE (0.926792-0.990458), junto con los valores más bajos de RMSE (0.101152-0.332819) y MAE (0.000386-0.034319). Los puntajes ponderados obtenidos confirmaron el dominio de ANFIS en múltiples estaciones, salvo en casos específicos: GEP se destacó en Bhubaneswar y M5P en Balangir. Este estudio resaltó la capacidad de ANFIS para predecir valores de ETo en diversas ubicaciones, lo cual se evidenció en la captura efectiva de variaciones en los diagramas de dispersión. Esto subraya la importancia de la selección de modelos, la versatilidad de los modelos de aprendizaje automático y la eficacia de combinar técnicas de inteligencia artificial (IA) para una predicción precisa de la ETo. ANFIS superó de manera consistente a M5P y GEP, consolidando su condición de herramienta confiable de predicción. Si bien se reconoce el potencial de M5P y GEP en contextos específicos, este estudio enfatiza la necesidad de adaptar los modelos a las características propias de cada lugar. Referencias a estudios relacionados respaldaron la eficacia de los enfoques hibridados de IA para mejorar el modelado de la ETo. Se invita a la investigación continua para refinar los modelos, incorporar factores adicionales y mejorar la precisión predictiva. Nuestros hallazgos ofrecen perspectivas valiosas para la gestión de los recursos hídricos, la planificación del riego y la toma de decisiones agrícolas en diversas ubicaciones.
References
[1] P. Aghelpour, V. Varshavian, M. Khodamorad Pour, and Z. Hamedi, “Comparing three types of data-driven models for monthly evapotranspiration prediction un-der heterogeneous climatic conditions,” Sci. Rep., vol. 12, no. 1, art. 22272 Dec. 2022. https://doi.org/10.1038/s41598-022-22272-3 DOI: https://doi.org/10.1038/s41598-022-22272-3
[2] D. K. Roy, R. Barzegar, J. Quilty, and J. Adamowski, “Using ensembles of adaptive neuro-fuzzy inference sys-tem and optimization algorithms to predict reference evapotranspiration in subtropical climatic zones,” J. Hy-drol., vol. 591, art. 125509, Dec. 2020. https://doi.org/10.1016/j.jhydrol.2020.125509 DOI: https://doi.org/10.1016/j.jhydrol.2020.125509
[3] H. Dehghanisanij, H. Emami, S. Emami, and V. Rezaver-dinejad, “A hybrid machine learning approach for esti-mating the water-use efficiency and yield in agriculture,” Sci. Rep., vol. 12, no. 1, art. 10844, Dec. 2022. https://doi.org/10.1038/s41598-022-10844-2 DOI: https://doi.org/10.1038/s41598-022-10844-2
[4] A. Elbeltagi et al., “Estimating the standardized precipi-tation evapotranspiration index using data-driven tech-niques: A regional study of Bangladesh,” Water, vol. 14, no. 11, art. 1764, Jun. 2022. https://doi.org/10.3390/w14111764 DOI: https://doi.org/10.3390/w14111764
[5] A. S. Azad et al., “Water level prediction through hy-brid SARIMA and ANN models based on time series analysis: Red Hills Reservoir case study,” Sustainability, vol. 14, no. 3, art. 1843, Feb. 2022. https://doi.org/doi: 10.3390/su14031843 DOI: https://doi.org/10.3390/su14031843
[6] A. Ashrafzadeh, O. Kişi, P. Aghelpour, S. M. Biazar, and M. A. Masouleh, “Comparative study of time series models, support vector machines, and gmdh in forecast-ing long-term evapotranspiration rates in northern Iran,” J. Irrig. Drain. Eng., vol. 146, no. 6, Jun. 2020, art. 04020005. https://doi.org/10.1061/(asce)ir.1943-4774.0001471 DOI: https://doi.org/10.1061/(ASCE)IR.1943-4774.0001471
[7] J. Wang et al., “Development of monthly reference evapotranspiration machine learning models and map-ping of Pakistan—A comparative study,” Water, vol. 14, no. 10, art. 1666, May 2022. https://doi.org/10.3390/w14101666 DOI: https://doi.org/10.3390/w14101666
[8] S. Mehdizadeh, B. Mohammadi, Q. B. Pham, and Z. Duan, “Development of boosted machine learning models for estimating daily reference evapotranspira-tion and comparison with empirical approaches,” Wa-ter, vol. 13, no. 24, art. 3489, Dec. 2021. https://doi.org/10.3390/w13243489 DOI: https://doi.org/10.3390/w13243489
[9] M. Poursaeid, A. H. Poursaeid, and S. Shabanlou, “A comparative study of artificial intelligence models and a statistical method for groundwater level prediction,” Water Resour. Manage., vol. 36, no. 5, pp. 1499–1519, Mar. 2022. https://doi.org/10.1007/s11269-022-03070-y DOI: https://doi.org/10.1007/s11269-022-03070-y
[10] R. M. Adnan, R. Mostafa, A. R. M. T. Islam, O. Kisi, A. Kuriqi, and S. Heddam, “Estimating reference evapo-transpiration using hybrid adaptive fuzzy inferencing coupled with heuristic algorithms,” Comput. Electron. Agric., vol. 191, Dec. 2021, art. 106541. https://doi.org/10.1016/j.compag.2021.106541 DOI: https://doi.org/10.1016/j.compag.2021.106541
[11] R. K. Jaiswal, A. K. Lohani, and R. V. Galkate, “Rainfall and agro related climate extremes for water require-ment in paddy grown Mahanadi Basin of India,” Agric. Res., vol. 12, no. 1, pp. 20–31, Mar. 2023. https://doi.org/10.1007/s40003-022-00629-4 DOI: https://doi.org/10.1007/s40003-022-00629-4
[12] S. Samantaray, A. Sahoo, and A. Agnihotri, “Assessment of flood frequency using statistical and hybrid neural network method: Mahanadi River Basin, India,” J. Geol. Soc. India, vol. 97, no. 8, pp. 867–880, Aug. 2021. https://doi.org/10.1007/s12594-021-1785-0 DOI: https://doi.org/10.1007/s12594-021-1785-0
[13] J. T. Abatzoglou, S. Z. Dobrowski, S. A. Parks, and K. C. Hegewisch, “TerraClimate, a high-resolution global dataset of monthly climate and climatic water balance from 1958-2015,” Sci. Data, vol. 5, art. 170191, Jan. 2018. https://doi.org/10.1038/sdata.2017.191 DOI: https://doi.org/10.1038/sdata.2017.191
[14] A. Choudhary, B. S. Das, K. Devi, and J. R. Khuntia, “ANFIS- and GEP-based model for prediction of scour depth around bridge pier in clear-water scouring and live-bed scouring conditions,” J. Hydroinfor., vol. 25, no. 3, pp. 1004–1028, May 2023. https://doi.org/10.2166/hydro.2023.212 DOI: https://doi.org/10.2166/hydro.2023.212
[15] T. Takagi and M. Sugeno, "Fuzzy identification of sys-tems and its applications to modeling and control," IEEE Tran. Syst. Man Cyber., vol. SMC-15, no. 1, pp. 116-132, Jan.-Feb. 1985. https://doi.org/10.1109/TSMC.1985.6313399 DOI: https://doi.org/10.1109/TSMC.1985.6313399
[16] C. Warren, “MATLAB for engineers: Development of an online, interactive, self-study course,” Eng. Educ., vol. 9, no. 1, pp. 86–93, 2014. https://doi.org/10.11120/ened.2014.00026 DOI: https://doi.org/10.11120/ened.2014.00026
[17] M. Lasheen and M. Abdel-Salam, “Maximum power point tracking using Hill Climbing and ANFIS techniques for PV applications: A review and a novel hybrid ap-proach,” Energy Convers. Manage., vol. 171, no. March, pp. 1002–1019, 2018. https://doi.org/10.1016/j.enconman.2018.06.003 DOI: https://doi.org/10.1016/j.enconman.2018.06.003
[18] R. Zhang and X. Xue, “A new model for prediction of soil thermal conductivity,” Int. Commun. Heat Mass Transf., vol. 129, art. 105661, Dec. 2021. https://doi.org/10.1016/j.icheatmasstransfer.2021.105661 DOI: https://doi.org/10.1016/j.icheatmasstransfer.2021.105661
[19] A. Adams and L. Sterling, Eds. AI ’92. Singapore: World Scientific, 1992. doi: 10.1142/1897 DOI: https://doi.org/10.1142/9789814536271
[20] M. Pal and S. Deswal, “M5 model tree based modeling of reference evapotranspiration,” Hydrol. Process., vol. 23, no. 10, pp. 1437–1443, May 2009. https://doi.org/10.1002/hyp.7266 DOI: https://doi.org/10.1002/hyp.7266
[21] M. Ali, R. C. Deo, N. J. Downs, and T. Maraseni, “An ensemble-ANFIS based uncertainty assessment model for forecasting multi-scalar standardized precipitation index,” Atmos. Res., vol. 207, art. 5771, November 2017, pp. 155–180, 2018. https://doi.org/10.1016/j.atmosres.2018.02.024 DOI: https://doi.org/10.1016/j.atmosres.2018.02.024
[22] P. Rai, P. Kumar, N. Al-Ansari, and A. Malik, “Evaluation of machine learning versus empirical models for month-ly reference evapotranspiration estimation in Uttar Pradesh and Uttarakhand States, India,” Sustainability, vol. 14, no. 10, art. 5771, May 2022. https://doi.org/10.3390/su14105771 DOI: https://doi.org/10.3390/su14105771
[23] M. Kadkhodazadeh, M. V. Anaraki, A. Morshed-Bozorgdel, and S. Farzin, “A new methodology for ref-erence evapotranspiration prediction and uncertainty analysis under climate change conditions based on ma-chine learning, multi criteria decision making and Monte Carlo methods,” Sustainability, vol. 14, no. 5, art. 2601, Mar. 2022. https://doi.org/10.3390/su14052601 DOI: https://doi.org/10.3390/su14052601
[24] D. K. Roy, A. Lal, K. K. Sarker, K. K. Saha, and B. Datta, “Optimization algorithms as training approaches for prediction of reference evapotranspiration using adap-tive neuro fuzzy inference system,” Agric. Water Manag., vol. 255, art. 107003, Sep. 2021. https://doi.org/10.1016/j.agwat.2021.107003 DOI: https://doi.org/10.1016/j.agwat.2021.107003
How to Cite
APA
ACM
ACS
ABNT
Chicago
Harvard
IEEE
MLA
Turabian
Vancouver
Download Citation
CrossRef Cited-by
Dimensions
PlumX
Article abstract page views
Downloads
License
Copyright (c) 2025 Deepak Raj, T Gopikrishnan
This work is licensed under a Creative Commons Attribution 4.0 International License.
The authors or holders of the copyright for each article hereby confer exclusive, limited and free authorization on the Universidad Nacional de Colombia's journal Ingeniería e Investigación concerning the aforementioned article which, once it has been evaluated and approved, will be submitted for publication, in line with the following items:
1. The version which has been corrected according to the evaluators' suggestions will be remitted and it will be made clear whether the aforementioned article is an unedited document regarding which the rights to be authorized are held and total responsibility will be assumed by the authors for the content of the work being submitted to Ingeniería e Investigación, the Universidad Nacional de Colombia and third-parties;
2. The authorization conferred on the journal will come into force from the date on which it is included in the respective volume and issue of Ingeniería e Investigación in the Open Journal Systems and on the journal's main page (https://revistas.unal.edu.co/index.php/ingeinv), as well as in different databases and indices in which the publication is indexed;
3. The authors authorize the Universidad Nacional de Colombia's journal Ingeniería e Investigación to publish the document in whatever required format (printed, digital, electronic or whatsoever known or yet to be discovered form) and authorize Ingeniería e Investigación to include the work in any indices and/or search engines deemed necessary for promoting its diffusion;
4. The authors accept that such authorization is given free of charge and they, therefore, waive any right to receive remuneration from the publication, distribution, public communication and any use whatsoever referred to in the terms of this authorization.
