Evaluation of the QUIC-URB wind solver and QESRadiant radiation-transfer model using a dense array of urban meteorological observations

This study assesses the performance of QUIC-URB, a fast-response urban flow model, and QESRadiant, a ray tracing radiation transfer model. Both models are components of the QUIC EnvSim 3D urban micro-scale model, which aims to simulate meteorological variables at high spatiotemporal resolution (~ 1 min, ~ 1 m) in urban settings. The evaluation was performed over a 5.85 ha sector of a university campus, in which complex 3D building geometry, vegetation, and various surface materials were modelled. First, wind speeds computed with QUIC-URB were compared to 30-min measurements over 10 days at 19 locations. Although results showed a significant underestimation for locations in the wake of buildings, considering model assumptions, its inexpensive computational cost, and measurement uncertainty, the agreement between computed and measured wind speeds is good (r² = 0.53, mean absolute error = 0.68 m s^{− 1}). Second, incoming radiation computed with QESRadiant was compared to 2-min measurements over seven clear sky days at 17 locations. Overall, the agreement between computed and measured incoming solar radiation was excellent (r² = 0.95). For both models, simulations were run successfully on a standard laptop machine with highly reasonable computational cost, on the order of minutes.