Computational Fluid Dynamics in Urban Air Quality Management

Computational Fluid Dynamics (CFD) has emerged as a vital tool for understanding and managing air quality in densely built environments. By numerically solving the governing equations of fluid flow and pollutant transport, CFD enables detailed characterisation of wind patterns, turbulence structures and scalar dispersion around buildings, street canyons and green infrastructure. These simulations inform the design of urban layouts, ventilation corridors and mitigation measures such as vegetation buffers or architectural modifications. Advances in turbulence modelling, high-performance computing and data‐driven surrogate models have expanded the scope of CFD from single‐street studies to city‐scale assessments, allowing planners to predict pollutant hotspots, evaluate the effectiveness of green walls and trees, and optimise urban form for ventilation. The integration of machine learning approaches with traditional CFD workflows is further reducing computational cost and facilitating rapid exploration of design alternatives, thereby supporting evidence‐based decision–making for healthier urban air.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Computational Fluid Dynamics in Urban Air Quality Management publication trend

The graph below shows the total number of articles in computational fluid dynamics in urban air quality management across all publications each year (not limited to Nature Index journals).

Technical terms

Computational Fluid Dynamics (CFD): The numerical solution of governing fluid equations to simulate flow, turbulence and scalar transport in complex geometries.

Reynolds-Averaged Navier–Stokes (RANS): A turbulence‐modelling approach that solves time‐averaged flow equations, reducing computational cost at the expense of detailed unsteady structures.

Large Eddy Simulation (LES): A high‐resolution technique that directly resolves large turbulent eddies while modelling smaller scales, offering improved accuracy for urban flow dynamics.

Street canyon: A configuration of parallel building façades forming a channel that influences wind flow and pollutant dispersion at pedestrian level.

Surrogate model: A data‐driven approximation, often machine learning-based, that replicates the behaviour of high‐fidelity simulations for rapid prediction and design optimisation.

References

1. Predicting Wind Comfort in an Urban Area: A Comparison of a Regression- with a Classification-CNN for General Wind Rose Statistics. Machine Learning and Knowledge Extraction (2024).
2. On the use of numerical modelling for near-field pollutant dispersion in urban environments − A review. Environmental Pollution (2015).
3. CFD and wind-tunnel analysis of outdoor ventilation in a real compact heterogeneous urban area: Evaluation using “air delay”. Building and Environment (2017).