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

For the effective planning and management of water resources and the implementation of related strategies, evaporation losses must be estimated properly, especially in regions that are prone to drought. Changes in climatic factors, such as in temperature, wind speed, sunshine hours, humidity, and solar radiation, can have a significant impact on the evaporation process. As such, evaporation is a highly non-linear, non-stationary process, and it can be difficult to model based on climatic factors, especially under different agroclimatic conditions. The aim of this study, therefore, is to investigate the applicability of several machine learning (ML) models (conditional random forest regression, multivariate adaptive regression splines, bagged multivariate adaptive regression splines, model tree M5, K-nearest neighbor, and weighted K-nearest neighbor) in modeling the monthly pan evaporation estimation. This study proposes the development of newly explored ML models for modeling evaporation losses in three different locations throughout Iraq based on the available climatic data. The evaluation of the performance of the proposed models based on various evaluation criteria showed the capability of the weighted K-nearest neighbor model for modeling the monthly evaporation losses in the study areas, with better accuracy when compared with the other existing models used as a benchmark in this study.

Automated summary

Evaporation losses must be properly estimated in regions prone to drought, as changes in climatic factors can significantly impact the evaporation process. This study explores the applicability of machine learning models for modeling monthly pan evaporation estimation in Iraq, with the weighted K-nearest neighbor model showing better accuracy compared to other existing models.