Daily Evapotranspiration Prediction at Arid and Semiarid Regions by Using Multiple Linear Regression Technique at Ramadi City in Iraq Region

Abstract

Abstract For many decades, because of the climate change, the interest in water resources management have been increased. The researchers curried out huge efforts to investigate the factors that influence the water resources quantities. Many of these efforts focused on studying the water losses by evapotranspiration. In the current study, a new mathematical model is introduced. It was built based on Multiple linear regression method (MLR). This model was utilized to estimate the daily evapotranspiration in Ramadi city which characterized with arid and semiarid environment and to investigate the influence of different parameters on the evapotranspiration process. The field data were collected from the digital meteorological station at the Upper Euphrates basin Developing Centre during the period from 23/11/2020 to 1/10/2022. These data include the evapotranspiration, maximum temperature, the minimum temperature, the average temperature, wind speed, relative humidity, and solar radiation. The model was calibrated using 80% of the data set, while 20% of the data set were used for validation. The solar radiation shows the highest impact on ET, while the lowest was recorded by relative humidity. High performance of the model was proved by testing it using Performance Indicators such as RMSE, NAE, MAPE, NSE, and R2.