Data-Driven Techniques for Monthly Pan Evaporation Modeling in Iraq

Tahsin, Ala (2020) Data-Driven Techniques for Monthly Pan Evaporation Modeling in Iraq. Eurasian Journal of Science & Engineering, 6 (1). pp. 104-120. ISSN 24145629

[img] Text (Research Article)
eajse.v6i1p104.pdf - Published Version
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (1MB)
Official URL:


Evaporation plays significant roles in agricultural production, climate change and water resources management. Hence, its accurate prediction is of paramount importance. This study aimed at investigating the potentials of artificial neural network (ANN), support vector regression (SVR) and classical multiple linear regression (MLR) models for monthly pan evaporation modeling in Erbil and Salahaddin stations of Iraq. Data including maximum, minimum, and mean temperatures, wind speed, relative humidity, and vapor pressure were used as inputs for 5 different input combinations to achieve the study objective. For performance evaluation of the applied models, root mean square error (RMSE) and determination coefficient (DC) were employed. In addition, Taylor diagrams were plotted to compare the performance of the models. The results showed that models with 6 inputs provided the best performance for Salahaddin station, but 5 inputs model led to better accuracy for MLR model in Erbil station. ANN provided superior performance with DC = 0.9527 and RMSE = 0.0660 for Erbil station while for Salahaddin station, SVR performed better with DC and RMSE of 0.8487 and 0.0753 in the validation phase. The general study results demonstrated that all the 3 applied models could be employed for successful pan evaporation modeling in the study stations, but for better accuracy, ANN is preferable.

Item Type: Article
Uncontrolled Keywords: Artificial Neural Network, Support Vector Regression, Erbil, Data, Salahaddin
Subjects: T Technology > TA Engineering (General). Civil engineering (General)
Depositing User: ePrints deposit
Date Deposited: 10 Feb 2021 06:51
Last Modified: 10 Feb 2021 06:51

Actions (login required)

View Item View Item