Introducing a Novel Double Hybrid Algorithm (DHA) and Developing Its Application for Predicting Air Temperature Under Climate Change Conditions (A Case Study of Iran Country)

Abstract

In the present study, a novel double hybrid algorithm (DHA) based on the least squares support vector machine (LSSVM) and hybrid Aquila optimization-particle swarm optimization (AO-PSO) algorithms, namely LSSVM-AO-PSO for minimum, mean, and maximum monthly air temperature (AT) prediction under climate change for periods (2020-2047) at 30 meteorological stations in Iran is presented. For this purpose, first, four benchmark data sets (BDSs) are used to prove the performance of the DHA. Then, modeling of minimum, mean and maximum AT in different time delays are performed, and the best time delay is selected. The technique for order of preference by similarity to the ideal solution (TOPSIS) method showed reasonable results of the DHA compared to LSSVM, LSSVM-PSO, and LSSVM-AO in modeling AT (in the best time delay). Results of changes in AT parameters under two models, ACCESS-ESM1 and CanESM5 global climate models (GCMs), and SSP scenarios in the future period (2020-2047) showed that AT was increased in all regions of Iran in a pessimistic state, and in an optimistic state were decreased in the northeast of Iran and were increased in other regions.