Validation of a Fast-Response Urban Microclimate Model Using a Dense Network of Weather Observations

Urban microclimate models are used in several applications, such as air quality control, energy efficiency in buildings or urban planning. For the latter, high spatial resolution is required since meteorological conditions can vary significantly over short distances mostly due to the spatial heterogeneity in surface thermal properties, building geometry, anthropogenic heat flux releases and vegetation cover.

This study presents a field validation of QUIC EnvSim (QES), a new-generation urban 3D model, fully scalable and physically-based, that enables energy, moisture and momentum exchanges between airflow and the underlying urban surface. QES combines QUIC, a CFD-based wind solver and dispersion model, and EnvSim, composed of a building-resolving radiation transfer model, a land-surface scheme and a turbulent transport model. Although some of the individual modules of QES have been previously validated, this study presents the first fully integrated validation of the modelling system. To do this, we used field measurements from the LUCE campaign, during which 92 meteorological stations were deployed over a sector of 0.1 km2 of the Swiss Institute of Technology in Lausanne (EPFL), from July 2006 to May 2007.

To assess the model performance, we ran simulations for several clear-sky days using 2-min time steps at a 1-m spatial resolution over a subset of the EPFL campus where most of the stations were located. Our analysis focuses on the ability of QES to simulate wind speed, incoming shortwave radiation, as well as surface and air temperatures throughout the urban canopy. So far, our validation analysis has shown that the QES radiation model performs very well at a relatively low computational cost. For instance, it can capture reasonably well (typically ± 30 min) the moment when a point on the surface becomes in the sun or in the shade due to solar obstruction by surrounding obstacles. Overall, the simulated incoming shortwave radiation is accurate when compared to in situ observations (R2 = 0.922), unless there are subtle surface geometry features (a small tree, a panel, etc.) that are not accounted for in the model. These preliminary results suggest that QES may be a valuable tool in decision-making regarding adaptation of urban planning.