PALM-4U

PALM-4U (“PALM for urban applications”) is the urban-focused component set of the PALM model system, a large-eddy simulation code developed for atmospheric boundary-layer research. PALM-4U adds what urban work needs on top of the LES core: building and plant canopy representation, a multi-layer surface energy balance for walls and roofs, radiative transfer that accounts for the sky view factor inside street canyons, and human biometeorology output.

The reason it exists is that the previous generation of urban climate models had to choose between resolving the geometry and covering a useful domain. Models that resolved buildings were confined to a few streets; models that covered a city parameterised the buildings away. PALM-4U targets the gap — building-resolving physics over a neighbourhood to city-district domain — which is the scale at which micro-scale processes actually determine outcomes and at which planning decisions are made.

What it resolves that a mesoscale model does not

• Flow around individual buildings, including wakes, channelling, and downwash into pedestrian level.
• Surface temperature separately for each wall, roof, and ground facet, rather than one effective value per grid cell.
• Radiative trapping inside canyons, where facing walls intercept longwave radiation that would otherwise escape — the mechanism behind much of the nocturnal urban heat island.
• Mean radiant temperature and the thermal indices that follow from it, which is what a heat-stress question actually needs and what a grid-cell air temperature cannot supply.

Practical notes

Output is netCDF, which puts it directly into the standard analysis stack and made the UC2 data standard work possible — the standard was developed in the same programme to make urban climate output from different groups comparable rather than merely available.

Two things are worth knowing before committing to it. The first is that runs are episodes, not climatologies: metre-scale resolution over a district costs enough that a study is a handful of characteristic days, and choosing those days well matters as much as configuring the physics. The second is that the input requirements are demanding — building geometry, surface materials, vegetation, and anthropogenic heat all have to come from somewhere, and in practice the quality of that static input often limits the result more than the model does.

See also: urban climate, urban atmosphere, large-eddy simulation, and the regional and urban climate modelling map.