Evaluation of M5P Algorithm for Estimation of Potential Evapotranspiration, Minimum and Maximum Temperature (Case study: Sari Weather Station)

Document Type : Original Article

Authors

1 Graduated ph.d, Dept. of Water Engineering, Faculty of Water and Soil, University of Zabol. Iran

2 Assistant Professor, Dept. of Water Engineering, Faculty of Agriculture, University of Zanjan, Iran

3 Master student of water resources engineering,Imam khomeini internatinal university, Iran

Abstract

In this study, we evaluated the performance of M5P tree model in estimating three parameters of potential evapotranspiration, minimum and maximum temperature. The used data of monthly mean of Sari Synoptic Weather Station during the 30 year period from 1989 to 2019 were performed. According to the results of M5P model with correlation coefficient of 0.92, RMSE of 0.25 and MAE of 0.12, has better performance in predicting potential transpiration evaporation than prediction of minimum and maximum temperature. . According to sensitivity analysis of M5P algorithm with average temperature inputs, sunshine hours, air humidity percentage, dew point temperature and saturation vapor pressure deficiency had the highest and with sunny hour inputs and wind speed had the least correct performance in predicting potential transpiration evaporation. It was also found that the wind velocity parameter had a negative effect on the prediction performance of the M5P model in the amount of transpiration evaporation.
In this study, we evaluated the performance of M5P tree model in estimating three parameters of potential evapotranspiration, minimum and maximum temperature. The used data of monthly mean of Sari Synoptic Weather Station during the 30 year period from 1989 to 2019 were performed. According to the results of M5P model with correlation coefficient of 0.92, RMSE of 0.25 and MAE of 0.12, has better performance in predicting potential transpiration evaporation than prediction of minimum and maximum temperature. . According to sensitivity analysis of M5P algorithm with average temperature inputs, sunshine hours, air humidity percentage, dew point temperature and saturation vapor pressure deficiency had the highest and with sunny hour inputs and wind speed had the least correct performance in predicting potential transpiration evaporation. It was also found that the wind velocity parameter had a negative effect on the prediction performance of the M5P model in the amount of transpiration evaporation.

Keywords

Journal of Meteorology and Atmospheric Science
Volume 2, Issue 4 - Serial Number 4
March 2020
Pages 287-295

Files

Share

How to cite

Statistics