Abstract: We assessed the solar photovoltaic power potential at Kunlun Station, Antarctica with the goal of establishing a photovoltaic plant at the station. We analyzed NASA satellite data of the station area from 2001 to 2017 and conducted forecasts of global irradiance. Our results show that the annual irradiance is high. During the year, there are monthly variations in the irradiance distribution, which remains relatively stable during each month. The average daily irradiance and sunshine hours also exhibit similar variations throughout the year. Using artificial neural networks, we conducted forecasts and climate feedback analysis. Results show that the irradiance is positively correlated with temperature, aerosol, and ozone content, while cloud amount and rainfall show negatively correlation with the irradiance in the Antarctic region.

Key words: Kunlun Station, irradiance, artificial neural network, Antarctica