The influence of data quality on urban heating demand modeling using 3D city models

• Published in: Computers, environment and urban systems Vol. 64; pp. 68 - 80
• Main Authors: Nouvel, Romain, Zirak, Maryam, Coors, Volker, Eicker, Ursula
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01-07-2017
• Subjects: 3D city model; CityGML; Data quality; Heating demand; LoD; Urban energy analysis; LoD; Heating demand; 3D city model; Urban energy analysis; CityGML; Data quality
• ISSN: 0198-9715; 1873-7587
• DOI: 10.1016/j.compenvurbsys.2016.12.005

3D city models are rich data sets for urban energy analyses, providing geometrical and semantic data required to estimate the energy demand of entire districts, cities and even regions. However, given the diverse availability, uncertainty and Level of Details of these data and the resources required to collect them, managing data quality is a common challenge of urban energy modeling. Knowing the influences of the different input data for different configurations and applications enables to control the result accuracy and recommend intelligent and adequate data collecting strategies, by assigning resources on the most important parameters. This paper investigates the influences of geometrical, meteorological, semantic and occupancy related data quality on the heating demand estimated by the urban energy simulation platform SimStadt, applied to the City of Ludwigsburg in Germany. A focus on a district with consumption data available at building block level allows for a critical comparison between estimated and measured energy demands. Although the quantified information presented in this paper is specific to a case study, the main trends and developed methods are transferrable to other urban energy analysis studies based on 3D city models.
•Sensitivity analyses for geometrical, semantic, occupancy-related, or meteorological input data have been realized.•Required urban data have been classified in three relevance categories i.e. Must-Have, Relevant-to-have and Nice-to-have.•The Must-have building attributes are the building function, year of construction, state of refurbishment and residence type.•Using 3D city models Level of Details 1 in comparison to Level of Details 2 leads to MAPE of 7.3% on the annual heating demands.