Anchored Branched Steady-state WInd Flow Transformer (AB-SWIFT): a metamodel for 3D atmospheric flow in urban environments
Please login to view abstract download link
Urban and industrial areas are vulnerable to accidental releases of pollutants. To accurately determine the pollutant's plume position and affected areas, it is essential to estimate the atmospheric flow around the affected site. This flow can be accurately computed using numerical methods of Computational Fluid Dynamics (CFD). However, CFD computation is costly and slow, making it unsuitable for emergency response. Deep learning surrogates offer a promising alternative as they are usually much faster. Atmospheric flows pose significant challenges for the application of deep learning models. Urban morphology and terrain geometry exert a strong influence on flow dynamics and can be hard to parametrize. Furthermore, meshes of realistic scenarios can reach tens of millions of nodes. Finally, meteorological conditions, and in particular the stability of the atmospheric stratification, play a critical role in the governing flow behavior and require careful consideration [2]. To tackle these challenges, we propose an Anchored Branched Steady-state WInd Flow Transformer (AB-SWIFT), a transformer-based model with an internal branched structure uniquely designed for atmospheric flow modeling. AB-SWIFT relies on the anchor attention mechanism [1], allowing scalability to hundreds of millions of mesh points. To the best of the author’s knowledge, AB-SWIFT is among the first works to apply transformer-type neural networks to atmospheric modeling. It also explicitly accounts for variable atmospheric stratification stability, which is typically neglected in existing models. We challenge our model on a specially designed database of atmospheric simulations with varying urban geometries and atmospheric stability conditions. Our model reaches the best accuracy on all predicted fields compared to the state of the art transformers and graph-based models.
