The development of evolutionary computing model for simulating reference evapotranspiration over Peninsular Malaysia

Reference evapotranspiration (ETo) is one of the foremost elements of the hydrology cycle which is essential for water resources management and irrigation applications. The current study is emphasized on the implementation of evolutionary computing models (i.e., gene expression programming (GEP)) for the simulation daily ETo in different locations of Peninsular Malaysia. The ETo models are developed using various input combinations of meteorological variables including air temperature (mean, maximum, and minimum), relative humidity, solar radiation, and mean wind speed. The in situ measurements of the ET are used to validate the model’s performance. The performance of the proposed GEP model is also compared with five well-established empirical formulations (EFs) developed based on the related climatological variability. The attained results evidenced the potential of GEP-derived ETo models in terms of all the statistical measures used. The best GEP model attained when all the meteorological variables are incorporated. However, the study revealed that the use of only temperature information can provide substantial predictability compared to EFs at all the studied stations across Peninsular Malaysia. This confirms the applicability of GEP in simulating ETo with fewer meteorological variables. The major advantage of GEP compared to other black box artificial intelligence algorithms is that GEP provides a set of equations which can be used by practitioners for reliable estimation of ETo at field with a fewer meteorological variable and, thus, can have wide applicability in water resources management.