Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2605.25679 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914599382548480 |
|---|---|
| author | Zhang, Yujia Qi, Jiaxi Chen, Ruiyan Liu, Yong Zhang, Yuzhou Kuang, Lyulin Zhang, Rita Cai, Shengze |
| author_facet | Zhang, Yujia Qi, Jiaxi Chen, Ruiyan Liu, Yong Zhang, Yuzhou Kuang, Lyulin Zhang, Rita Cai, Shengze |
| contents | Accurate prediction of three-dimensional (3D) wind fields over complex mountainous terrain is essential for renewable energy deployment and regional weather modeling. Traditional computational fluid dynamics (CFD) simulations face two fundamental bottlenecks: expert-intensive mesh generation around irregular topography, and iterative solvers that require hours to days even on high-performance clusters. Recent neural operator approaches accelerate inference, but typically fail to resolve the sharp, localized velocity gradients induced by complex terrain features. Here, we present a transformer-based dual-attention neural-operator framework for 3D wind field prediction over complex mountainous terrain, and validate its effectiveness through two instantiations on representative point-based (mesh-free) and graph-based neural-operator architectures, namely Patch-solver and Patch-GTO. Trained on a large CFD-generated dataset spanning diverse terrain geometries and inflow conditions, the framework enables rapid prediction of steady-state wind field while maintaining competitive accuracy. It also demonstrates robust zero-shot transfer to real-world mountainous sites across several diverse locations, outperforming existing neural operator baselines by 10% in relative error. We further verify that incorporating sparse observational data (1% spatial coverage) reduces prediction error by 16.89% relative to the corresponding model without sparse data input and by 32.75% relative to advanced neural operator baselines on unseen terrains. This framework establishes a generalizable computational paradigm across domains, promising to be a real-time tool for wind resource assessment over complex mountainous terrain and related atmosphere-surface interaction studies. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_25679 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Transformer-based Neural Operators for 3D Wind Field Prediction over Complex Mountainous Terrain Zhang, Yujia Qi, Jiaxi Chen, Ruiyan Liu, Yong Zhang, Yuzhou Kuang, Lyulin Zhang, Rita Cai, Shengze Fluid Dynamics Accurate prediction of three-dimensional (3D) wind fields over complex mountainous terrain is essential for renewable energy deployment and regional weather modeling. Traditional computational fluid dynamics (CFD) simulations face two fundamental bottlenecks: expert-intensive mesh generation around irregular topography, and iterative solvers that require hours to days even on high-performance clusters. Recent neural operator approaches accelerate inference, but typically fail to resolve the sharp, localized velocity gradients induced by complex terrain features. Here, we present a transformer-based dual-attention neural-operator framework for 3D wind field prediction over complex mountainous terrain, and validate its effectiveness through two instantiations on representative point-based (mesh-free) and graph-based neural-operator architectures, namely Patch-solver and Patch-GTO. Trained on a large CFD-generated dataset spanning diverse terrain geometries and inflow conditions, the framework enables rapid prediction of steady-state wind field while maintaining competitive accuracy. It also demonstrates robust zero-shot transfer to real-world mountainous sites across several diverse locations, outperforming existing neural operator baselines by 10% in relative error. We further verify that incorporating sparse observational data (1% spatial coverage) reduces prediction error by 16.89% relative to the corresponding model without sparse data input and by 32.75% relative to advanced neural operator baselines on unseen terrains. This framework establishes a generalizable computational paradigm across domains, promising to be a real-time tool for wind resource assessment over complex mountainous terrain and related atmosphere-surface interaction studies. |
| title | Transformer-based Neural Operators for 3D Wind Field Prediction over Complex Mountainous Terrain |
| topic | Fluid Dynamics |
| url | https://arxiv.org/abs/2605.25679 |