Recent advances in black box and white-box models for urban heat island prediction: Implications of fusing the two methods

The urban heat island (UHI) phenomenon is a serious concern for urban planners and policymakers, requiring effective and efficient mitigation policies. To develop such policies, accurate and pre-emptive estimations of current and future UHI manifestations are vital elements that help determine efficient policies and mitigation techniques. There are two fundamental approaches for modelling overheating in an urban environment: whitebox and black-box based methods. The first one is characterized by the easily interpretable working process, while the unclear working procedure defines the second one. The present study comprehensively reviews the commonly used white-box and black-box based approaches applied for UHI predictions, analyses the existing literature adopting these tools for UHI prediction, and discusses the effectiveness of fusing both methods at the design and operation stages of the urban area for effective prediction and mitigation of UHI effect. The literature analysis showed that the transparent working process and high prediction accuracy of the physical-based whitebox models make them a popular and reliable tool for UHI evaluation. Nevertheless, some white-box based simulation tools are too complex and require a high level of expertise to operate, leading to potential inaccuracies in the obtained outcomes. Black-box models, in turn, despite their opaque working process, are more straightforward in use and require less computation time. The fusion of these two methods is a novel approach that may benefit both UHI prediction and mitigation at the design and operation stages, respectively.

Automated summary

The whitebox models are popular due to their transparent working process and high prediction accuracy, but some simulation tools are too complex and require a high level of expertise to operate, potentially leading to inaccuracies in the outcomes. On the other hand, black-box models, despite their opaque working process, are easier to use and require less computation time. The fusion of these two methods is a novel approach that may benefit both UHI prediction and mitigation at the design and operation stages.