The implications of sky model selection for the prediction of daylight distribution in architectural spaces

This paper investigates the implications of the selection of various sky luminance distribution models for the computational prediction of daylight distribution in architectural spaces. The illuminance distribution in an actual test-space is simulated based on six different sky models, and the results are compared with illuminance measurements taken in the test space. The variations of simulation results and their relationship with the measurements are presented and discussed. PROBLEM STATEMENT In a separate study, the predictions of six sky luminance models were compared with actual luminance measurements obtained with a sky scanner in a tropical context (Lam et al. 1997). This paper explores the implications of the sky model selection for the prediction of illuminance distribution inside architectural spaces. Specifically, we seek to answer the following questions: a) What variations in the simulated illuminance levels for a typical architectural space is to be expected as a result of using different sky luminance models? b)Is the extent of this divergence significant in view of other uncertainties involved in the design and evaluation process (e.g., building description, validity of the daylight simulation program, etc.)? c) What can be said of the relative performance of various sky luminance models given the available measurement results