Estimation of Reference Evapotranspiration Using Neural Networks and Cuckoo Search Algorithm

Shahaboddin Shamshirband, Mohsen Amirmojahedi, Milan Gocić, Shatirah Akib, Dalibor Petković, Jamshid Piri, Slavisa Trajkovic

Research output: Contribution to journalArticleResearchpeer-review

Abstract

The ability to optimize an artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS) in reference evapotranspiration (ET0) estimation using the cuckoo search algorithm (CSA) is studied in this paper. The monthly series of climatic data (minimum and maximum air temperatures, actual vapor pressure, sunshine hours, and wind speed at height of 2.0 m) from twelve meteorological stations in Serbia during the period 1983–2010 were used as inputs to the soft computing models. As the reference ET0 equation, the FAO-56 Penman-Monteith equation was selected. Statistical indicators such as the root-mean-square error (RMSE), mean absolute error (MAE), and coefficient of determination (R2) were used as comparing criteria for the evaluation of the models’ performances. The obtained results show that the proposed ANFIS+CSA model can be used for ET0 estimation with high reliability (RMSE=0.2650  mm day−1, MAE=0.1843 and R2=0.9695). The selected soft computing models were compared with the results of two empirical models (adjusted Hargreaves and Priestley-Taylor) and their calibrated versions. Priestley-Taylor method had the highest RMSE (0.5420  mm day−1). The lowest RMSE of 0.1883  mm day−1 has the ANN model. The calibrated adjusted Hargreaves model performs better than the calibrated Priestley-Taylor model. The ANN+CSA, ANFIS, and ANFIS+CSA had better characteristics than the two estimated empirical equations and their calibrated versions.
Original languageEnglish
Pages (from-to)-
JournalJournal of Irrigation and Drainage Engineering
DOIs
Publication statusPublished - 18 Sep 2015

Fingerprint

Evapotranspiration
neural networks
evapotranspiration
Neural networks
Fuzzy inference
Mean square error
artificial neural network
Serbia
Vapor Pressure
Soft computing
Neural Networks (Computer)
Sunlight
Air
Penman-Monteith equation
Temperature
Food and Agricultural Organization
Adaptive algorithms
vapor pressure
Vapor pressure
wind speed

Cite this

Shamshirband, Shahaboddin ; Amirmojahedi, Mohsen ; Gocić, Milan ; Akib, Shatirah ; Petković, Dalibor ; Piri, Jamshid ; Trajkovic, Slavisa. / Estimation of Reference Evapotranspiration Using Neural Networks and Cuckoo Search Algorithm. In: Journal of Irrigation and Drainage Engineering. 2015 ; pp. -.
@article{2ef0e53be96849b38ca95e0a4d935b03,
title = "Estimation of Reference Evapotranspiration Using Neural Networks and Cuckoo Search Algorithm",
abstract = "The ability to optimize an artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS) in reference evapotranspiration (ET0) estimation using the cuckoo search algorithm (CSA) is studied in this paper. The monthly series of climatic data (minimum and maximum air temperatures, actual vapor pressure, sunshine hours, and wind speed at height of 2.0 m) from twelve meteorological stations in Serbia during the period 1983–2010 were used as inputs to the soft computing models. As the reference ET0 equation, the FAO-56 Penman-Monteith equation was selected. Statistical indicators such as the root-mean-square error (RMSE), mean absolute error (MAE), and coefficient of determination (R2) were used as comparing criteria for the evaluation of the models’ performances. The obtained results show that the proposed ANFIS+CSA model can be used for ET0 estimation with high reliability (RMSE=0.2650  mm day−1, MAE=0.1843 and R2=0.9695). The selected soft computing models were compared with the results of two empirical models (adjusted Hargreaves and Priestley-Taylor) and their calibrated versions. Priestley-Taylor method had the highest RMSE (0.5420  mm day−1). The lowest RMSE of 0.1883  mm day−1 has the ANN model. The calibrated adjusted Hargreaves model performs better than the calibrated Priestley-Taylor model. The ANN+CSA, ANFIS, and ANFIS+CSA had better characteristics than the two estimated empirical equations and their calibrated versions.",
author = "Shahaboddin Shamshirband and Mohsen Amirmojahedi and Milan Gocić and Shatirah Akib and Dalibor Petković and Jamshid Piri and Slavisa Trajkovic",
year = "2015",
month = "9",
day = "18",
doi = "10.1061/(ASCE)IR.1943-4774.0000949",
language = "English",
pages = "--",
journal = "Journal of Irrigation and Drainage Engineering - ASCE",
issn = "1943-4774",
publisher = "American Society of Civil Engineers (ASCE)",

}

Estimation of Reference Evapotranspiration Using Neural Networks and Cuckoo Search Algorithm. / Shamshirband, Shahaboddin; Amirmojahedi, Mohsen; Gocić, Milan; Akib, Shatirah; Petković, Dalibor; Piri, Jamshid; Trajkovic, Slavisa.

In: Journal of Irrigation and Drainage Engineering, 18.09.2015, p. -.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Estimation of Reference Evapotranspiration Using Neural Networks and Cuckoo Search Algorithm

AU - Shamshirband, Shahaboddin

AU - Amirmojahedi, Mohsen

AU - Gocić, Milan

AU - Akib, Shatirah

AU - Petković, Dalibor

AU - Piri, Jamshid

AU - Trajkovic, Slavisa

PY - 2015/9/18

Y1 - 2015/9/18

N2 - The ability to optimize an artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS) in reference evapotranspiration (ET0) estimation using the cuckoo search algorithm (CSA) is studied in this paper. The monthly series of climatic data (minimum and maximum air temperatures, actual vapor pressure, sunshine hours, and wind speed at height of 2.0 m) from twelve meteorological stations in Serbia during the period 1983–2010 were used as inputs to the soft computing models. As the reference ET0 equation, the FAO-56 Penman-Monteith equation was selected. Statistical indicators such as the root-mean-square error (RMSE), mean absolute error (MAE), and coefficient of determination (R2) were used as comparing criteria for the evaluation of the models’ performances. The obtained results show that the proposed ANFIS+CSA model can be used for ET0 estimation with high reliability (RMSE=0.2650  mm day−1, MAE=0.1843 and R2=0.9695). The selected soft computing models were compared with the results of two empirical models (adjusted Hargreaves and Priestley-Taylor) and their calibrated versions. Priestley-Taylor method had the highest RMSE (0.5420  mm day−1). The lowest RMSE of 0.1883  mm day−1 has the ANN model. The calibrated adjusted Hargreaves model performs better than the calibrated Priestley-Taylor model. The ANN+CSA, ANFIS, and ANFIS+CSA had better characteristics than the two estimated empirical equations and their calibrated versions.

AB - The ability to optimize an artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS) in reference evapotranspiration (ET0) estimation using the cuckoo search algorithm (CSA) is studied in this paper. The monthly series of climatic data (minimum and maximum air temperatures, actual vapor pressure, sunshine hours, and wind speed at height of 2.0 m) from twelve meteorological stations in Serbia during the period 1983–2010 were used as inputs to the soft computing models. As the reference ET0 equation, the FAO-56 Penman-Monteith equation was selected. Statistical indicators such as the root-mean-square error (RMSE), mean absolute error (MAE), and coefficient of determination (R2) were used as comparing criteria for the evaluation of the models’ performances. The obtained results show that the proposed ANFIS+CSA model can be used for ET0 estimation with high reliability (RMSE=0.2650  mm day−1, MAE=0.1843 and R2=0.9695). The selected soft computing models were compared with the results of two empirical models (adjusted Hargreaves and Priestley-Taylor) and their calibrated versions. Priestley-Taylor method had the highest RMSE (0.5420  mm day−1). The lowest RMSE of 0.1883  mm day−1 has the ANN model. The calibrated adjusted Hargreaves model performs better than the calibrated Priestley-Taylor model. The ANN+CSA, ANFIS, and ANFIS+CSA had better characteristics than the two estimated empirical equations and their calibrated versions.

U2 - 10.1061/(ASCE)IR.1943-4774.0000949

DO - 10.1061/(ASCE)IR.1943-4774.0000949

M3 - Article

SP - -

JO - Journal of Irrigation and Drainage Engineering - ASCE

JF - Journal of Irrigation and Drainage Engineering - ASCE

SN - 1943-4774

ER -