Comparative Performance Analysis of Artificial Bee Colony and Particle Swarm Optimization Algorithms in Modeling Pan Evaporation

Abstract

Evaporation phenomenon is one of the most difficult parameters to predict in the hydrological cycle. It is considered as an important variable in determining reservoir capacities and irrigation water needs, in water budget calculations, and in the formation of heat flow between the earth and the atmosphere. In this respect, successful modeling of evaporation phenomenon is an important issue in many respects. Many methods such as artificial intelligence methods and machine learning techniques have been used in evaporation modeling in the literature. In recent years, metaheuristic optimization algorithms, which were inspired by the foraging behavior of living creatures in nature, have also begun to be applied in hydrological phenomena. In the current study, pan evaporation modeling was carried out with the help of Artificial Bee Colony and Particle Swarm Optimization algorithms, which are commonly used metaheuristic methods, and the performances of both methods were examined comparatively. In the studies carried out, monthly data between 2002 and 2019 of the meteorology station number 17230 which is belong to Turkish State Meteorological Service, located in Anamur district in the south of Turkey were used. In the evaluation of the results, Coefficient of Determination (R2), Nash-Sutcliffe Efficiency Coefficient (NSE) and Mean Square Error (MSE) values were used as performance criteria. The results obtained at the end of the study were examined numerically with performance indicators and visually with time series graphs and Taylor diagrams.