Abstract
Natural daylight is recognized as an important variable in the energy performance of buildings. A method that estimates the global illuminance received on a horizontal surface at ground level and its direct component at normal incidence under cloudless conditions is presented. The method uses the k-distribution method and the correlated-k approximation to compute a set of clearness indices integrated over 13 spectral bands covering the range 380–780 nm. A spectral resampling technique, including a spectral disaggregation and a spectral linear interpolation, is applied to these indices for providing a detailed set of solar irradiances at 1 nm in spectral resolution over the whole range. Then, these are weighted by the standardized CIE action spectrum for human eye for assessing the illuminance. Inputs to the method include the total column contents of ozone and water vapor as well as aerosol optical properties produced by the Copernicus Atmosphere Monitoring Service. Estimates of illuminance were compared to high-quality 1 min measurements of illuminance that were collected from two experimental sites located in two different climatic zones. A slight overestimation is observed for the global illuminance: the bias is between +1 klx and +3 klx, i.e., between +1% and +4% in relative value. The root mean square error varies between 5 klx (8%) and 6 klx (9%). The squared correlation coefficient ranges between 0.95 and 0.97. At the site providing the direct illuminance at normal incidence, the performance of the method is lower compared to global illuminance with a lower squared correlation coefficient of 0.53. The bias, relative bias, RMSE, and rRMSE are +7 klx, +9%, 12 klx, and 15%, respectively. The uncertainty of the method is of the order of the uncertainty of the measurements. The method offers accurate estimates of illuminance in cloudless conditions at high spatial and temporal resolutions useful for construction industries and operators as well as thermal simulation tools for optimal building design strategies.
Highlights
Natural daylight radiation reaching the surface of the Earth has become an essential climatic variable in the human daily life
This paper describes the method to derive the illuminance from the clearness indices in the 13 KBs, which are themselves calculated from runs of libRadtran with inputs from Copernicus Atmosphere Monitoring Service (CAMS)
Despite data being available at two stations only, it is worth investigating the sources of errors resulting in an overall overestimation by the presented method: slightly positive bias on the global illuminance at horizontal surface and higher positive bias on direct illuminance at normal incidence at Vaux-en-Velin
Summary
Natural daylight radiation reaching the surface of the Earth has become an essential climatic variable in the human daily life. The McClear clear-sky model [13,14] is an example of the operational use of the Kato et al [12] scheme combined with inputs from the Copernicus Atmosphere Monitoring Service (CAMS) for computing total irradiances under cloudless conditions. It accurately reproduces the irradiance computed by libRadtran in a much faster way by using pre-computed abaci, which are known as look-up tables, for each of the 32 spectral bands and the band 240–4606 nm, combined with interpolation functions. This study contributes to answering the growing need of construction industries and investors on the spatial and temporal availability of illuminance at ground level for optimal building design strategies
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
