view:18443 Last Update: 2021-3-1
تخمين تبخير و تعرق مرجع در شرايط داده محدود با استفاده از شبکه هاي عصبي RBF
Estimation of reference crop evapotranspiration in limited weather data condition using RBF neural networks
Reference crop evapotranspiration is one of the important factors of hydrological cycle. This parameter is used to design irrigation systems, hydraulic structures and drainage systems. Penman- Monteith FAO-56 equation is a standard method to estimate reference crop evapotranspiration, however it needs many meteorological data. This meteorological data is not available at all synoptic stations. In this study, the capabilities of RBF neural networks in the condition of complete weather data and limited weather data has been investigated at Zanjan synoptic station. Totally 24 scenarios of different weather data has been investigated. Results showed that the RBF neural network can estimate the reference crop evapotranspiration with high precision. If the data of temperature exist, the model can predict the evapotranspiration with the maximum RMSE=0.6 mm/day. The results of different scenarios showed that, combination of maximum temperature and wind speed gives the best results when the weather data is limited.