FLUME-FNO uses a Fourier neural operator with a new Multi-Directional Distance Feature to predict 3D urban wind and temperature fields from limited CFD data on unseen building morphologies, reporting 0.2 m/s and 0.19°C mean absolute errors.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
physics.flu-dyn 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
FLUME-FNO: data-efficient and scalable prediction of 3D wind and temperature fields in unseen urban morphologies
FLUME-FNO uses a Fourier neural operator with a new Multi-Directional Distance Feature to predict 3D urban wind and temperature fields from limited CFD data on unseen building morphologies, reporting 0.2 m/s and 0.19°C mean absolute errors.