{"references": ["Grobe and Jakubiec (in press). Impact of model detail on daylighting metrics in residential buildings. Proceedings CISBAT 2023.", "Forouzandeh (2023). Influence of inaccurate material optical properties and geometrical levels of detail on daylight simulation results - sensitivity analysis and uncertainty quantification. 21st International Radiance Workshop. Innsbruck, Austria. Aug. 2023", "McNeil (2014). BSDFs, matrices and phases. 13th International Radiance Workshop. London, UK, Sept. 2014.", "Ward et al. (2021). Modeling specular transmission of complex fenestration systems with data-driven BSDFs. Building and Environment, 196, 107774. doi: 10.1016/j.buildenv.2021.107774", "Wasilewski et al. (in press). Raytraverse: Navigating the lightfield to enhance climate-based daylight modeling. In: Proceedings SimAUD 2021.", "Standfest et al. (2022). Swiss dwellings: A large dataset of apartment models including aggregated geolocation-based simulation results covering viewshed, natural light, traffic noise, centrality and geometric analysis. Zenodo. doi: 10.5281/zenodo.7070951", "Jakubiec et al. (2019). Subjective and measured evidence for residential lighting metrics in the tropics. In: Proceedings Building Simulation Conference 2019. 2.", "Daylight in buildings. Standard EN 17037:2018. CEN / TC 169.", "Approved method: IES Spatial Daylight Autonomy (SDA) and Annual Sunlight Exposure (ASE). Recommendation LM-83-12. IES."]}

Climate-Based Daylight Modelling (CBDM) differs from other daylight simulation techniques in that it can produce representative results in physical and photometric units. The typical application of CBDM during the design of buildings or the planning of interventions however implies incomplete information, e.g. on the definitive furnishing, selection of surface properties, and installation of shades and glare controls. This is reflected by uncertainties and reduced level of detail (LoD) in the setup of models. Relying on a detailed model of an existing residential unit in Singapore, we assess the impact LoD by sequentially reducing model detail. Impact is assessed in terms of a) absolute quantities, e.g. illuminance, b) thresholded metrics, e.g. Daylight Autonomy and Daylight Glare Probability, and c) relative ranking based on these metrics. We demonstrate the implementation of the efficient zonal daylight simulation with Raytraverse and Radiance, and the processing of the spatio-temporal results with Python (and its numpy and scipy modules). An outlook into the possible extension of the study by accessing a building database is provided.

Supported by Velux Stiftung.