Statistical modelling of Urban Heat Island using Local Climate Zone classification.

Air temperature over cities shows a very strong spatial and temporal heterogeneity, and therefore, models to properly predict this variability are often complex and computationaly heavy.

Yet, it is sometimes difficult to establish the strongly needed dialogue with policy makers and city planners when models are too complex to comprehend, or when a long time is needed to produce simulations of different scenarios and their impact on societally relevant issues, like human thermal comfort, air quality etc.

Predicting how intense the urban heat island intensity can be during summer night conditions, and where it is most likely to be strong, using only input data that stakeholders can comprehend is probably an effective first step.

Relatively simple methods are proposed with this contribution, which can approximate the results of far more heavy thermal physics models with an acceptable residual mean square error and a good spatial representation. They combine factorial analysis and linear models with mixed effects, using Local Climate Zones, population, and accessible geographical indicators as predictors.

The loss of precision is a good trade-off in regard of the gain in explainability and rapidity of use. Once the parameters of the model are estimated, one can explore the impact of a major urban renovation project with almost no delay as long as the geographical information before and after the project are available.