Modeling the Thermal Impact of a Highway on Its Surroundings Using PALM: A Case Study in the Atlantic Rainforest, Brazil

The near-surface air temperature (Ta) is largely driven by surface temperature (TS), which depends on the surface energy balance and is strongly influenced by surface material properties. While TS can be estimated globally from satellite observations at high spatial resolution, Ta is typically measured only at weather stations, resulting in limited spatial coverage. Consequently, many studies use TS in place of Ta when a high spatial resolution is needed (e.g., at the city scale), although their spatial patterns differ and their physical relationships are not fully characterized. Previous studies using Landsat-derived TS (30 m resolution) have shown that surface changes can affect the surrounding thermal environment, with significant TS anomalies extending up to 168 m into adjacent undisturbed areas. One notable case is a highway constructed in 2013 across the dense Atlantic Rainforest in southeastern Brazil, where TS increases extended asymmetrically into the forest, reaching up to 128 m on one side of the road and 64 m on the other. This reach and asymmetry were hypothesized to result from warm air advection from the highway into the forest canopy, modulated by the prevailing sea-breeze circulation. In this study, we use the PALM large-eddy simulation model to isolate and test this hypothesis by explicitly simulating the Atlantic Rainforest with and without the highway. Idealized simulations are performed under a mean wind speeds perpendicular to the road axis to assess how the highway modifies air temperature at canopy height and how these effects propagate into the forest on both sides. The simulations show an increase in Ta of more than 1.0°C reaching 50m inside the forest canopy and of 0.5°C reaching 100m. The reach can occur on both sides and are not just driven by the mean wind but also by the increase in soil and leafs exposure to the sun radiation because of vegetation removal. The results provide a mechanistic understanding of how this type of infrastructure can alter the microclimate of protected forested areas and help clarify the relationship between TS-based observations and Ta responses, also providing guidance for the use of Ts when Ta observations are unavailable.