Impact of atmospheric components on solar clear-sky models at different elevation: Case study Canary Islands

• Assessment on the performance of solar clear-sky models at different altitude.
• SOLIS and REST2 clear-sky models were superior with fine atmospheric inputs.
• ESRA proved more robust with low spatial resolution atmospheric inputs.
• Over-estimation occurred at the lower site when using inputs from the upper site.

The estimation of clear-sky solar irradiance via clear-sky models depends on reliable values of aerosol optical depth, water vapor and ozone content. These atmospheric variables are rarely on-site measured and are generally provided as gridded estimates in very low spatial resolution (1°). The high spatial variability of atmospheric variables within the grid resolution (pixel) leads to important errors in those areas with great atmospheric variability, such as in mountainous regions. In this paper, the performance of three clear-sky solar irradiance models was evaluated in a site with especially great elevation range, the Izana station from the Baseline Surface Radiation Network (Tenerife, Canary Islands) located at a high elevation (2373 m) and just 14 km from the ocean. Aerosols data were obtained from measurements from the Aerosol Robotic Network (AERONET) at the same site. The evaluation was also compared with global horizontal irradiance estimations with clear-sky models in the Guimar station, located at a lower elevation (156 m) and only 11.5 km away from Izana. Results showed a strong influence of elevation on solar radiation estimation under clear-sky conditions.