Investigating the performance of Kstar and GPR Algorithms for modeling meteorological drought index RDI (case study: east of Urmia lake basin)

Authors

1 Water Eng. Urmia Unbversity

2 Water Eng. Urmia University

3 Department of Hydrology and Water Resource, University of Shaheed Chamran, Ahvaz, Iran

Abstract

Drought is one of the natural calamities that causes many losses to different societies every year. Drought monitoring can provide water resource managers and planners with valuable information to develop plans to deal with drought and reduce its related damages. In the present research, the meteorological drought of four synoptic stations of Ahar, Jolfa, Tabriz and Maragheh in East Azarbaijan province, east of Lake Urmia, was investigated and analyzed using the RDI drought index in the statistical period of 1955 to 2019. To calculate the RDI index, monthly potential evaportranspiration and precipitation data were used with the FAO Penman-Monteith method. The RDI index was calculated for each station separately and for three time scales of 6, 9 and 12 months, and the results of the calculations showed that drought monitoring is better than other time scales in the 12-month time scale. Then, using Kstar and GPR algorithms, the RDI index was modeled in three time scales. The results of the evaluation criteria showed the high performance of the mentioned algorithms, so that the numerical value of the correlation coefficient of the GPR algorithm for all stations is 0.92 and the same coefficient with the Kstar algorithm is in the range of 0.91 to 0.92. Also, the numerical value of RMSE with the GPR algorithm was between 0.31 and 0.39 and Kstar algorithm was 0.32 and 0.51. From these results, it can be claimed that the GPR algorithm has modeled the RDI drought index with higher accuracy.

Keywords

Main Subjects

Journal of Civil and Environmental Engineering

Articles in Press, Accepted Manuscript
Available Online from 17 May 2023

Files

Share

How to cite

Statistics