🔥 Deep learning has been widely explored almost in all research filed. Here is a curated list of papers about deep learning methods in Earth System, especially relating to weather prediction. It also contains frameworks for data-driven Numerical Weather Prediciton (NWP) training, tools to deploy weather prediction, courses and tutorials about Ai4earth and all publicly available weather prediction checkpoints and APIs.
- Awesome-ai4earth
-
- Extreme Weather and Prediction
tropical cyclones, heat wave etc.
- Climate phenomena analysis
- Extreme Weather and Prediction
- [2023-11-07] Creat this project, add some big model for atmospherical modeling!
- Add more paper, datasets, research directions for ai4earth ✨Contributions Wanted
Also check out the project that I am currently working on: EarthVision - A deep learniong framwork for Numerical Weather Prediction, Earth System Data cmpression, Precipitation Prediction)
| Date | keywords | Institute | Paper | Publication |
|---|---|---|---|---|
| 2022-02 | FourCastNet | Nvidia | FourCastNet: Accelerating Global High-Resolution Weather Forecasting Using Adaptive Fourier Neural Operators [paper1][paper2] | PASC  |
| 2022-11 | PanguWeather | HuaWei | Accurate medium-range global weather forecasting with 3D neural networks [paper1][paper2] | Nature |
| 2022-12 | GraphCast | DeepMind | GraphCast: Learning skillful medium-range global weather forecasting | Science |
| 2023-04 | FengWu | Shanghai AILab | FengWu: Pushing the Skillful Global Medium-range Weather Forecast beyond 10 Days Lead | arxiv |
| 2023-06 | FuXi | Fudan University | FuXi: A cascade machine learning forecasting system for 15-day global weather forecast | npj  |
- FengWu-Adas Towards an End-to-End Artificial Intelligence Driven Global Weather Forecasting System. (arxiv, 12/2023)
- FengWu-4DVar FengWu-4DVar: Coupling the Data-driven Weather Forecasting Model with 4D Variational Assimilation (arxiv, 12/2023)
- NeuralGCM Neural General Circulation Models. (arxiv, 11/2023)
- GenCast Diffusion-based ensemble forecasting for medium-range weather. (arxiv. 12/2023)
- MetNet-3 Deep Learning for Day Forecasts from Sparse Observations. (arxiv, 06/2023)
- ClimaX A foundation model for weather and climate, (ICML, 2023) [project]
- NowcastNet Skilful nowcasting of extreme precipitation with NowcastNet. (nature, 07/2023)
- PostRainBench - A comprehensive benchmark and a new model for precipitation forecasting. (arxiv, 10/2023)
- Anthropogenic fingerprints in daily precipitation revealed by deep learning. (nature, 08/2023)
- CorrDiff Generative residual diffusion modeling for km-scale atmospheric downscaling, (arxiv, 09/2023)
- Two deep learning-based bias-correction pathways improve summer precipitation prediction over China, (Environmental Research Letters, 12/2022) paper
- FuXi-Extreme: Improving extreme rainfall and wind forecasts with diffusion mode. (arxiv, 10/2023)
- The Potential Predictability of Fire Danger Provided by Numerical Weather Prediction (JAMC, 11/2016)
- Machine learning–based observation-constrained projections reveal elevated global socioeconomic risks from wildfire. (nature communication, 22/03/2023)
The following list makes sure that all weather forecasting models are compared apples to apples.
WeatherBench 2: A benchmark for the next generation of data-driven global weather models [project] [LeaderBoard] (08/2023)
You may also find these leaderboards helpful:
- WeatherBench2 Leaderboard - evaluating and comparing various weather forecasting models. displays up-to-date scores of many of the state-of-the-art ML and physics-based models.
If you're interested in the field of AI4Earth, you may find the above list of milestone papers helpful to explore its history and state-of-the-art. However, each direction of AI4Earth offers a unique set of insights and contributions, which are essential to understanding the field as a whole. For a detailed list of papers in various subfields, please refer to the following link (it is possible that there are overlaps between different subfields):
(:exclamation: We would greatly appreciate and welcome your contribution to the following list. âť—)
-
Analyse different NWP models in different fields with respect to different abilities
-
some deeplearning methods in wildfire forecast.
- Era5 single-level - ERA5 is the fifth generation ECMWF reanalysis for the global climate and weather for the past 8 decades. Data is available from 1940 onwards.
- Era5 pressure-level - ERA5 is the fifth generation ECMWF reanalysis for the global climate and weather for the past 8 decades. Data is available from 1940 onwards.
- Era5-Land - ERA5-Land is a reanalysis dataset providing a consistent view of the evolution of land variables over several decades at an enhanced resolution compared to ERA5
| Model | Size | Architecture | Access | Date | Origin | Model License1 |
|---|---|---|---|---|---|---|
| PanguWeather | ~740M | 3D Transformer | ckpt | 2022-11 | Paper | BY-NC-SA 4.0 |
| GraphCast | unkonwn | GNN | ckpt | 2022-12 | Paper | Apache 2.0 & BY-NC-SA 4.0 |
| FengWu | unkonwn | U-Transformer | ckpt | 2023-06 | Paper | - |
| FuXi | unkonwn | Transformer | ckpt | 2023-04 | Paper | - |
- DeepSpeed - DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.
- Megatron-DeepSpeed - DeepSpeed version of NVIDIA's Megatron-LM that adds additional support for several features such as MoE model training, Curriculum Learning, 3D Parallelism, and others.
- FairScale - FairScale is a PyTorch extension library for high performance and large scale training.
- Megatron-LM - Ongoing research training transformer models at scale.
- Colossal-AI - Making large AI models cheaper, faster, and more accessible.
- BMTrain - Efficient Training for Big Models.
- Mesh Tensorflow - Mesh TensorFlow: Model Parallelism Made Easier.
- maxtext - A simple, performant and scalable Jax LLM!
- Alpa - Alpa is a system for training and serving large-scale neural networks.
- GPT-NeoX - An implementation of model parallel autoregressive transformers on GPUs, based on the DeepSpeed library.
- Fluid-Earth - An interactive web application that allows you to visualize current and past conditions of Earth's atmosphere and oceans.
- TODO - A curated (still actively updated) list of practical guide resources of earth-related research.
This is an active repository and your contributions are always welcome!
I will keep some pull requests open if I'm not sure if they are awesome for ai4earth, you could vote for them by adding đź‘Ť to them.
If you have any question about this opinionated list, do not hesitate to contact me hantao10200@gmail.com.
Footnotes
-
This is not legal advice. Please contact the original authors of the models for more information. ↩