Abstract
This research evaluated the spatiotemporal dimension of the thermal environment of urban microclimate system in Singapore through a case study. To provide a holistic assessment of air temperature due to urban morphology, for the spatial dimension, correlations between different urban morphology variables and estate-level air temperature were examined through machine learning models to test nonlinear assumption. Different urban morphologies were measured and calculated including shading coefficient of buildings, shading coefficient of greenery, green plot ratio, building height, building cover ratio, building wall area, impervious ground surface fraction, and ground albedo. The temporal dimension of air temperature changes was examined at hourly resolution. The test score of hourly temperature predictions using nonlinear algorithm ranged from 0.24 °C to 0.67 °C, with RMSE above 0.5 °C occurring at T8-T17. Furthermore, this research aggregated hourly temperature prediction into an average cumulative temperature increase (ACTI). A heatmap was created showing ACTI on a university campus with a maximum temperature difference of 2 K between cold and hot spots. Based on the investigation, this study further proposed guidelines for low carbon urban planning for greenery, painting or pavement material, and active system.
Published Version
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