On the use of UAV-thermal imaging for CFD validation of urban thermal microclimate

Computational Fluid Dynamics (CFD) simulations are widely applied to assess urban microclimate processes and adaptation strategies. However, their accuracy and reliability critically depend on high-quality validation data. Although field measurements and reduced-scale wind-tunnel testing are commonly used for validation, the use of remote sensing methods remains limited despite their potential. This study systematically evaluates UAV-based thermal imaging as a tool for validating microscale CFD predictions of urban thermal conditions. The case study focuses on the city centre of Semarang, Indonesia, where ground-truth field measurements and UAV surveys with visible RGB and infrared (IR) imaging are conducted. CFD simulations are performed for the same area using 3D URANS with the realizable k–ε model on a high-resolution computational grid. Results show that (i) UAV-IR data and ground-truth measurements differed by 0.47 °C – 1.40 °C across asphalt, concrete, and soil–grass surfaces ranged from, confirming the accuracy of UAV thermal imaging; and (ii) CFD simulations deviated by 3.88 °C in surface-averaged temperatures for impervious-based areas compared with UAV-IR data. These findings highlight UAV thermal imaging as a practical and data-driven approach for validating CFD models, enabling more robust analyses and design of sustainable urban thermal environments.