Collection of refined architectural parameters by crowdsourcing using Facebook social network: Case of Greater Tunis

Abstract The urban heat island, especially in low-income countries, increases the consequences of heatwaves. Urban micro-climate studies could help to manage heatwave crises and improve urban planning. However, they require accurate urban data, and such are often not available in African countries. This paper presents a crowdsourcing methodology using a very popular social media in this region, Facebook, to gather architectural information on buildings in the Greater Tunis agglomeration. Following WUDAPT protocol, the questions encompassed number of floors, window pattern, color, wall fabric, air-conditioning (AC) and building's use. More than 100 answers were gathered. An expert validation shows an accuracy of more than 85% on all parameters. The geo-localized sampled buildings were associated to Local Climate Zones. These are shown to be pertinent in Tunis. In this hot and dry climate, recent houses with many windows are found to be more equipped with AC than those using traditional construction methods with just a few windows. Mid-rise buildings follow modern construction paradigm, with many or even plate glass windows, and therefore use AC. Finally, this application of crowdsourcing to urban climate science through a social media questionnaire allows to propose architectural parameters for urban climate models in this region of North Africa.