• Regional Climate Models
• Local-Impact Models
  • MEGAN-MOHYCAN
  • REGCROP
  • SURFEX
  • UrbClim
  • WAM and COHERENS
• Domains and spatial resolution
• Dynamical regional climate simulations
• Statistical downscaling
• Status simulations

Urban model: UrbClim

UrbClim is a three-dimensional urban climate model, developed at VITO (De Ridder et al., 2015). The model is composed of a boundary layer model, which is coupled to a land surface scheme containing urban physics. Typically covering an entire urban agglomeration (domain size of a few tens of kilometers) at a spatial resolution of a few hundred meters, the model accounts for the effect of large-scale atmospheric circulations through the specification of external forcing fields provided by global or regional climate models, focusing on the simulation of urban climate as a deviation from the large-scale fields.

Compared to full mesoscale models, UrbClim makes a series of simplifying assumptions. Among other things, the model does not calculate an internal pressure gradient, nor does it represent atmospheric radiation or cloud physics; instead, all these parameters are specified as external fields from the forcing (host) model. As a result, the model is faster by several orders of magnitude compared to traditional mesoscale models operating at the same resolution, so that it can simulate fairly long periods, covering tens of years, and at the same time considering an ensemble of external forcing models, as well as several climate and land cover change scenarios.

In the context of on-going Belgian and European research projects, the UrbClim model has been extensively tested for several European cities, including Antwerp, Gent, Brussels, Bilbao, Almada, Berlin, Skopje, and London. Despite the simplifications made in the model, it consistently shows a good capacity to reproduce observed climate variables. In particular, the model captures well the observed urban-rural temperature differences, i.e., the urban heat island intensity, reproducing the timing as well as the magnitude of the observed peak values.