Prediction of evaporation using chaos theory and artificial intelligence in dry lands(case study: Semnan Province)

Articles in Press

Document Type : Original Article

Authors

1 PhD Graduated of Semnan University

2 Associate professor-Semnan University

3 Professor, Faculty of Civil Engineering, Semnan University, Semnan, Iran

10.22034/iwrr.2024.426028.2721

Abstract

Evaporation is one of the important phenomena of the hydrological cycle and its prediction is essential in water management, planning and conservation. Since chaos theory deals with the study of dynamic systems, therefore, in this research, the prediction of the evaporation process was carried out using the combination of chaos theory and intelligent models, including support vector machine, decision tree, group learning, and Gaussian process.ِData of the Semnan synoptic station was selected during the period of 1995-2019. The optimal values of delay and mutual information were obtained using false nearest neighbor methods in order to reconstruct the variable phase space of evaporation, equal to 18 and 9, respectively. According to different combinations of variables, the most optimal response of all models was determined for the combination of all parameters, and the two factors of evaporation and temperature had the greatest impact on the prediction. In general, the support vector machine model with R2 = 85.5 and MAE = 1.4 had the best performance, and then the methods of Gaussian process, group learning and decision tree method performed well. The combined use of chaos theory along with intelligent algorithms has a good ability to estimate evaporation.

Keywords

Main Subjects

Iran-Water Resources Research

Articles in Press, Accepted Manuscript
Available Online from 23 April 2024

Files

History

  • Receive Date: 19 November 2023
  • Revise Date: 18 April 2024
  • Accept Date: 23 April 2024

Share

How to cite

Statistics