Extreme Learning Machines: A new approach for prediction of reference evapotranspiration

Shafika Sultan Abdullah, Marlinda Abdul Malek, Namiq Sultan Abdullah, Ozgur Kisi, Keem Siah Yap

Research output: Contribution to journalArticle

71 Citations (Scopus)

Abstract

Recognizing the importance of precise determination of reference evapotranspiration (ET0) is a principal step in the attempts to reserve huge quantities of squandered water. This paper investigates the efficiency of Extreme Learning Machines (ELM) algorithm at predicting Penman-Monteith (P-M) ET0 for Mosul, Baghdad, and Basrah meteorological stations, located at the north, mid, and southern part of Iraq. Data of weather parameters containing maximum air temperature (Tmax), minimum air temperature (Tmin), sunshine hours (Rn), relative humidity (Rh), and wind speed (U2) for the period (2000-2013) are used as inputs to the ELM model by using four different input cases including complete and incomplete sets of meteorological data. The performance of ELM model is compared with the empirical P-M equation and with feedforward backpropagation (FFBP) model. The evaluation criteria used for comparison are the root of mean squared error (RMSE), mean absolute error (MAE), and coefficient of determination (R2). The statistical results of both models are found to be encouraging; particularly results of running the ELM model with incomplete sets of data, noticing that the sensitivity of the proposed model to missing data changes from one location to another, as well as along the year for certain study location. The Rn is found to be the most effective parameter in Mosul Station, while U2 and Rh are found to act almost in parallel with Rn in Baghdad Station, and for conditions of Basrah Station; U2 and Rh prove to be the dominant parameters. The minimum and maximum time intervals required for running ELM model for all stations, and in all applied conditions, are (4.64-6.19) seconds respectively, while the same order of timing required for running the FFBP model is (6.30-27.80) seconds. The maximum R2 recorded for the ELM model is 0.991, while for the FFBP it is 0.985. The ELM proved to be efficient, simple in application, of high speed, and has very good generalization performance; therefore, this algorithm is highly recommended for locations similar to the geographical and meteorological conditions of Iraq that consists of both arid and semiarid regions.

Original languageEnglish
Pages (from-to)184-195
Number of pages12
JournalJournal of Hydrology
Volume527
DOIs
Publication statusPublished - 01 Aug 2015

Fingerprint

evapotranspiration
prediction
relative humidity
air temperature
machine learning
Penman-Monteith equation
semiarid region
arid region
wind velocity
station
weather
parameter

All Science Journal Classification (ASJC) codes

  • Water Science and Technology

Cite this

Abdullah, Shafika Sultan ; Abdul Malek, Marlinda ; Abdullah, Namiq Sultan ; Kisi, Ozgur ; Yap, Keem Siah. / Extreme Learning Machines : A new approach for prediction of reference evapotranspiration. In: Journal of Hydrology. 2015 ; Vol. 527. pp. 184-195.
@article{f9c04057aaba420f8ad6b6707fd77adb,
title = "Extreme Learning Machines: A new approach for prediction of reference evapotranspiration",
abstract = "Recognizing the importance of precise determination of reference evapotranspiration (ET0) is a principal step in the attempts to reserve huge quantities of squandered water. This paper investigates the efficiency of Extreme Learning Machines (ELM) algorithm at predicting Penman-Monteith (P-M) ET0 for Mosul, Baghdad, and Basrah meteorological stations, located at the north, mid, and southern part of Iraq. Data of weather parameters containing maximum air temperature (Tmax), minimum air temperature (Tmin), sunshine hours (Rn), relative humidity (Rh), and wind speed (U2) for the period (2000-2013) are used as inputs to the ELM model by using four different input cases including complete and incomplete sets of meteorological data. The performance of ELM model is compared with the empirical P-M equation and with feedforward backpropagation (FFBP) model. The evaluation criteria used for comparison are the root of mean squared error (RMSE), mean absolute error (MAE), and coefficient of determination (R2). The statistical results of both models are found to be encouraging; particularly results of running the ELM model with incomplete sets of data, noticing that the sensitivity of the proposed model to missing data changes from one location to another, as well as along the year for certain study location. The Rn is found to be the most effective parameter in Mosul Station, while U2 and Rh are found to act almost in parallel with Rn in Baghdad Station, and for conditions of Basrah Station; U2 and Rh prove to be the dominant parameters. The minimum and maximum time intervals required for running ELM model for all stations, and in all applied conditions, are (4.64-6.19) seconds respectively, while the same order of timing required for running the FFBP model is (6.30-27.80) seconds. The maximum R2 recorded for the ELM model is 0.991, while for the FFBP it is 0.985. The ELM proved to be efficient, simple in application, of high speed, and has very good generalization performance; therefore, this algorithm is highly recommended for locations similar to the geographical and meteorological conditions of Iraq that consists of both arid and semiarid regions.",
author = "Abdullah, {Shafika Sultan} and {Abdul Malek}, Marlinda and Abdullah, {Namiq Sultan} and Ozgur Kisi and Yap, {Keem Siah}",
year = "2015",
month = "8",
day = "1",
doi = "10.1016/j.jhydrol.2015.04.073",
language = "English",
volume = "527",
pages = "184--195",
journal = "Journal of Hydrology",
issn = "0022-1694",
publisher = "Elsevier",

}

Extreme Learning Machines : A new approach for prediction of reference evapotranspiration. / Abdullah, Shafika Sultan; Abdul Malek, Marlinda; Abdullah, Namiq Sultan; Kisi, Ozgur; Yap, Keem Siah.

In: Journal of Hydrology, Vol. 527, 01.08.2015, p. 184-195.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Extreme Learning Machines

T2 - A new approach for prediction of reference evapotranspiration

AU - Abdullah, Shafika Sultan

AU - Abdul Malek, Marlinda

AU - Abdullah, Namiq Sultan

AU - Kisi, Ozgur

AU - Yap, Keem Siah

PY - 2015/8/1

Y1 - 2015/8/1

N2 - Recognizing the importance of precise determination of reference evapotranspiration (ET0) is a principal step in the attempts to reserve huge quantities of squandered water. This paper investigates the efficiency of Extreme Learning Machines (ELM) algorithm at predicting Penman-Monteith (P-M) ET0 for Mosul, Baghdad, and Basrah meteorological stations, located at the north, mid, and southern part of Iraq. Data of weather parameters containing maximum air temperature (Tmax), minimum air temperature (Tmin), sunshine hours (Rn), relative humidity (Rh), and wind speed (U2) for the period (2000-2013) are used as inputs to the ELM model by using four different input cases including complete and incomplete sets of meteorological data. The performance of ELM model is compared with the empirical P-M equation and with feedforward backpropagation (FFBP) model. The evaluation criteria used for comparison are the root of mean squared error (RMSE), mean absolute error (MAE), and coefficient of determination (R2). The statistical results of both models are found to be encouraging; particularly results of running the ELM model with incomplete sets of data, noticing that the sensitivity of the proposed model to missing data changes from one location to another, as well as along the year for certain study location. The Rn is found to be the most effective parameter in Mosul Station, while U2 and Rh are found to act almost in parallel with Rn in Baghdad Station, and for conditions of Basrah Station; U2 and Rh prove to be the dominant parameters. The minimum and maximum time intervals required for running ELM model for all stations, and in all applied conditions, are (4.64-6.19) seconds respectively, while the same order of timing required for running the FFBP model is (6.30-27.80) seconds. The maximum R2 recorded for the ELM model is 0.991, while for the FFBP it is 0.985. The ELM proved to be efficient, simple in application, of high speed, and has very good generalization performance; therefore, this algorithm is highly recommended for locations similar to the geographical and meteorological conditions of Iraq that consists of both arid and semiarid regions.

AB - Recognizing the importance of precise determination of reference evapotranspiration (ET0) is a principal step in the attempts to reserve huge quantities of squandered water. This paper investigates the efficiency of Extreme Learning Machines (ELM) algorithm at predicting Penman-Monteith (P-M) ET0 for Mosul, Baghdad, and Basrah meteorological stations, located at the north, mid, and southern part of Iraq. Data of weather parameters containing maximum air temperature (Tmax), minimum air temperature (Tmin), sunshine hours (Rn), relative humidity (Rh), and wind speed (U2) for the period (2000-2013) are used as inputs to the ELM model by using four different input cases including complete and incomplete sets of meteorological data. The performance of ELM model is compared with the empirical P-M equation and with feedforward backpropagation (FFBP) model. The evaluation criteria used for comparison are the root of mean squared error (RMSE), mean absolute error (MAE), and coefficient of determination (R2). The statistical results of both models are found to be encouraging; particularly results of running the ELM model with incomplete sets of data, noticing that the sensitivity of the proposed model to missing data changes from one location to another, as well as along the year for certain study location. The Rn is found to be the most effective parameter in Mosul Station, while U2 and Rh are found to act almost in parallel with Rn in Baghdad Station, and for conditions of Basrah Station; U2 and Rh prove to be the dominant parameters. The minimum and maximum time intervals required for running ELM model for all stations, and in all applied conditions, are (4.64-6.19) seconds respectively, while the same order of timing required for running the FFBP model is (6.30-27.80) seconds. The maximum R2 recorded for the ELM model is 0.991, while for the FFBP it is 0.985. The ELM proved to be efficient, simple in application, of high speed, and has very good generalization performance; therefore, this algorithm is highly recommended for locations similar to the geographical and meteorological conditions of Iraq that consists of both arid and semiarid regions.

UR - http://www.scopus.com/inward/record.url?scp=84929299317&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84929299317&partnerID=8YFLogxK

U2 - 10.1016/j.jhydrol.2015.04.073

DO - 10.1016/j.jhydrol.2015.04.073

M3 - Article

AN - SCOPUS:84929299317

VL - 527

SP - 184

EP - 195

JO - Journal of Hydrology

JF - Journal of Hydrology

SN - 0022-1694

ER -