Modeling of pan evaporation based on the development of machine learning methods

For effective planning and management of water resources implementation the related strategies, it is important to ensure proper estimation evaporation losses, especially in regions that are prone drought. Changes climatic factors, such as changes temperature, wind speed, sunshine hours, humidity, solar radiation can have a significant impact on process. As such, highly non-linear, non-stationary process, be difficult modeled based different agro-climatic conditions. The aim this study, therefore, investigate feasibility several machines learning (ML) models (conditional random forest regression, Multivariate Adaptive Regression Splines, Bagged Model Tree M5, K- nearest neighbor, weighted neighbor) for modeling monthly pan estimation. This study proposes development newly explored ML losses three locations over Iraq region available data areas. evaluation performance proposed model various criteria showed capability neighbor studies areas with better accuracy when compared other existing used benchmark study.