مقايسه برآورد تابش خورشيد با استفاده از روش هارگريوز- ساماني و شبكه هاي عصبي مصنوعي

Solar Radiation Estimation Using Temperature-based and Artificial Neural Networks approaches

Radiation is a meteorological variable that governs many hydrological and phonological processes, but its measurements are not made routinely. To overcome this problem, continuous hydrological models that include evapotranspiration, snowmelt (using solar radiation data) and plant growth modules have applied different strategies to generate daily radiation data. In this paper, artificial neural networks (ANNs) and temperature-based (TB) approaches for estimation of solar radiation was used and compared. These two approaches were assessed by the records of Tehran-Mehrabad. Shiraz, Zahedan and Ammameh Stations. Results revealed better performance of ANNs than TB. However, the TB method because of its capability to generalize results and to be easily linked with hydrological models appears to be a good candidate to be applied in the catchments where the climatological data are limited.