Chaining the DART and SOLENE-microclimat models to support the use of TIR satellite data in urban climate studies

Land surface temperature (LST) derived from thermal infrared (TIR) satellite data is widely used in urban climate research due to the repetitive availability of data over large areas worldwide. LST directly reflects the interactions between urban surfaces, the atmosphere and human activities, supporting hotspots identification, comfort indices estimation or mitigation strategies planning. However, its use is limited by the spatial and temporal resolutions of current spaceborne sensors. The upcoming TIR satellite missions (LSTM, TRISHNA, SBG), with spatial resolution between 37 and 60 m and up to 3-days revisit, open up new opportunities to study urban climate at the neighborhood scale. At this scale, retrieving accurate and comparable LST over cities remains a challenge. Urban heterogeneity and 3D structure greatly impact satellite measurements, requiring a good understanding of 3D radiative processes for reliable LST estimates. Another challenge is the transition to air temperature, which is essential for improving comfort and quality of life in cities.

To address these challenges, a model chaining approach is implemented to generate physically coherent datasets linking remote sensing measurements to microclimate variables over any urban configuration in order to investigate how they relate to each other. On the one hand, the DART radiative transfer model simulates radiative exchanges in the urban canopy and the corresponding remotely sensed images, provided that the surface temperature distribution in the 3D urban scene is known. On the other hand, the thermo-radiative model SOLENE-microclimat simulates the surface temperature distribution in the 3D scene required by DART, as well as the air temperature in the canopy but does not allow the simulation of multispectral satellite data. Chaining the two models bridges the gap between remotely sensed TIR parameters and microclimate variables. This presentation gives an overview of the modelling chain and presents some concrete examples of its application to urban climate studies.