Lune

AAAI2024Top-tier venue

Earthfarsser: Versatile Spatio-Temporal Dynamical Systems Modeling in One Model

Hao Wu, Yuxuan Liang, Wei Xiong, Zhengyang Zhou, Wei Huang, Shilong Wang, Kun Wang

Abstract

Efficiently modeling spatio-temporal (ST) physical processes and observations presents a challenging problem for the deep learning community. Many recent studies have concentrated on meticulously reconciling various advantages, leading to designed models that are neither simple nor practical. To address this issue, this paper presents a systematic study on existing shortcomings faced by off-the-shelf models, including lack of local fidelity, poor prediction performance over long time-steps, low scalability, and inefficiency. To systematically address the aforementioned problems, we propose an EarthFarseer, a concise framework that combines parallel local convolutions and global Fourier-based transformer architectures, enabling dynamically capture the local-global spatial interactions and dependencies. EarthFarseer also incorporates a multi-scale fully convolutional and Fourier architectures to efficiently and effectively capture the temporal evolution. Our proposal demonstrates strong adaptability across various tasks and datasets, with fast convergence and better local fidelity in long time-steps predictions. Extensive experiments and visualizations over eight human society physical and natural physical datasets demonstrates the state-of-the-art performance of EarthFarseer. We release our code at https://github.com/easylearningscores/EarthFarseer.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

13 top-tier papers have cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • PURE: Prompt Evolution with Graph ODE for Out-of-distribution Fluid Dynamics Modeling cites An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
  • PURE: Prompt Evolution with Graph ODE for Out-of-distribution Fluid Dynamics Modeling cites Swin Transformer: Hierarchical Vision Transformer using Shifted Windows
  • Causal Deciphering and Inpainting in Spatio-Temporal Dynamics via Diffusion Model cites An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
  • Causal Deciphering and Inpainting in Spatio-Temporal Dynamics via Diffusion Model cites Earthformer: Exploring Space-Time Transformers for Earth System Forecasting
  • Causal Deciphering and Inpainting in Spatio-Temporal Dynamics via Diffusion Model cites SimVP: Simpler yet Better Video Prediction
  • An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale cites Scaling Autoregressive Video Models
  • Earthformer: Exploring Space-Time Transformers for Earth System Forecasting cites Scaling Autoregressive Video Models
  • Earthformer: Exploring Space-Time Transformers for Earth System Forecasting cites An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
  • Earthformer: Exploring Space-Time Transformers for Earth System Forecasting cites Swin Transformer: Hierarchical Vision Transformer using Shifted Windows
  • Swin Transformer: Hierarchical Vision Transformer using Shifted Windows cites An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
  • SimVP: Simpler yet Better Video Prediction cites Scaling Autoregressive Video Models
  • SimVP: Simpler yet Better Video Prediction cites An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
  • SimVP: Simpler yet Better Video Prediction cites Swin Transformer: Hierarchical Vision Transformer using Shifted Windows
  • Frequency-Aligned Knowledge Distillation for Lightweight Spatiotemporal Forecasting cites PURE: Prompt Evolution with Graph ODE for Out-of-distribution Fluid Dynamics Modeling
  • Frequency-Aligned Knowledge Distillation for Lightweight Spatiotemporal Forecasting cites SimVP: Simpler yet Better Video Prediction
  • FlowCast: Advancing Precipitation Nowcasting with Conditional Flow Matching cites Earthformer: Exploring Space-Time Transformers for Earth System Forecasting
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers13

  1. PURE: Prompt Evolution with Graph ODE for Out-of-distribution Fluid Dynamics ModelingHao Wu, Changhu Wang, Fan Xu, Jinbao Xue et al.NeurIPS 2024 · 23 citations
  2. Frequency-Aligned Knowledge Distillation for Lightweight Spatiotemporal ForecastingYuqi Li, Chuanguang Yang, Hansheng Zeng, Zeyu Dong et al.ICCV 2025 · 23 citations
  3. Causal Deciphering and Inpainting in Spatio-Temporal Dynamics via Diffusion ModelYifan Duan, Jian Zhao, pengcheng, Junyuan Mao et al.NeurIPS 2024 · 14 citations
  4. Improving Generalization of Dynamic Graph Learning via Environment PromptKuo Yang, Zhengyang Zhou, Qihe Huang, Limin Li et al.NeurIPS 2024 · 14 citations
  5. FlowCast: Advancing Precipitation Nowcasting with Conditional Flow MatchingBernardo Perrone Ribeiro, Jana Faganeli PucerICLR 2026 · 9 citations

Builds on15

  1. Swin Transformer: Hierarchical Vision Transformer using Shifted WindowsZe Liu, Yutong Lin, Yue Cao, Han Hu et al.ICCV 2021 · 31,683 citations
  2. An Image is Worth 16x16 Words: Transformers for Image Recognition at ScaleAlexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn et al.ICLR 2021 · 21,477 citations
  3. Earthformer: Exploring Space-Time Transformers for Earth System ForecastingZhihan Gao, Xingjian Shi, Hao Wang, Yi Zhu et al.NeurIPS 2022 · 410 citations
  4. SimVP: Simpler yet Better Video PredictionZhangyang Gao, Cheng Tan, Lirong Wu, Stan Z. LiCVPR 2022 · 313 citations
  5. Scaling Autoregressive Video ModelsDirk Weissenborn, Oscar Täckström, Jakob UszkoreitICLR 2020 · 252 citations

Related papers

  1. CSTDFormer: Empowering Transformers to Learn Spatio-Temporal Delays via Contrastive LearningEn Wang, Jiajian Lv, Di Liang, Zidie Zhou et al.INFOCOM 2026
  2. Sparse-Scale Transformer with Bidirectional Awareness for Time Series ForecastingYing Liu, Bo Liu, Sheng Huang, Gang Luo et al.AAAI 2026
  3. FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series ForecastingTian Zhou, Ziqing Ma, Qingsong Wen, Xue Wang et al.ICML 2022 · 2,912 citations
  4. InParformer: Evolutionary Decomposition Transformers with Interactive Parallel Attention for Long-Term Time Series ForecastingHaizhou Cao, Zhenhao Huang, Tiechui Yao, Jue Wang et al.AAAI 2023 · 29 citations
  5. PESD-TSF: A Period-Aware and Explicit Structured Decomposition Framework for Long-Term Time Series ForecastingHua Wang, Xianhao Jiao, Fan ZhangICML 2026
Dusk over the sea between two cliffs drawn in fine vertical lines