FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operators

By Jaideep Pathak, Shashank Subramanian, P. Harrington, S. Raja, A. Chattopadhyay, Morteza Mardani, Thorsten Kurth, D. Hall, Zong-Yi Li, Kamyar Azizzadenesheli and 3 more. arXiv.org, 2022

1,246 citations recorded by the source corpus.

FourCastNet, short for Fourier Forecasting Neural Network, is a global data-driven weather forecasting model that provides accurate short to medium-range global predictions at $0.25^{\circ}$ resolution. FourCastNet accurately forecasts high-resolution, fast-timescale variables such as the surface wind speed, precipitation, and atmospheric water vapor. It has important implications for planning wind energy resources, predicting extreme weather events such as tropical cyclones, extra-tropical cyclones, and atmospheric rivers. FourCastNet matches the forecasting accuracy of the ECMWF Integrated Forecasting System (IFS), a state-of-the-art Numerical Weather Prediction (NWP) model, at short lead times for large-scale variables, while outperforming IFS for variables with complex fine-scale structure, including precipitation. FourCastNet generates a week-long forecast in less than 2 seconds, orders of magnitude faster than IFS. The speed of FourCastNet enables the creation of rapid and inexpensive large-ensemble forecasts with thousands of ensemble-members for improving probabilistic forecasting. We discuss how data-driven deep learning models such as FourCastNet are a valuable addition to the meteorology toolkit to aid and augment NWP models.

Subfields: Atmospheric Science, Electrical and Electronic Engineering, Environmental Engineering

View Paper: https://www.semanticscholar.org/paper/10194e9d1d6b8ca8870445c990d4933c1dac1125

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Atmospheric Science

In Electrical and Electronic Engineering

In Environmental Engineering

Citations in the corpus

The operational medium-range deterministic weather forecasting can be extended beyond a 10-day lead time (2023, positive): “The combination of deterministic FengWu and FengWu-Ensemble (FengWu-ENS) effectively meets the needs of medium-range forecasting, allowing users to obtain various outcomes based on their speci ﬁ c requirements, including smoothed large-scale atmospheric circulation and detailed extreme events.”
ClimaX: A foundation model for weather and climate (2023, positive): “One can employ a multi-step objective for finetuning as in Pathak et al. (2022) to achieve better results.”
FourCastNet: Accelerating Global High-Resolution Weather Forecasting Using Adaptive Fourier Neural Operators (2022, background): “[22] that the IFS ensemble forecast consumes 2.”
WeatherBench 2: A Benchmark for the Next Generation of Data‐Driven Global Weather Models (2023, methodology): “Pathak et al. (2022) used a modified vision transformer (Guibas et al., 2022), called FourCastNet, for prediction at a very high resolution of 0.25° and a 6 hr time step.”
Geometry-Informed Neural Operator for Large-Scale 3D PDEs (2023, methodology): “In the 3D UNet model, we evaluate channel dimensions ranging from [64, 128, 256] and depths from [4, 5, 6].”
Earthformer: Exploring Space-Time Transformers for Earth System Forecasting (2022, methodology): “Pathak et al. [28] proposed the FourCastNet for global weather forecasting, which is based on Adaptive Fourier Neural Operators (AFNO) [14].”
GraphCast: Learning skillful medium-range global weather forecasting (2022, methodology): “Performance of GraphCast compared to top ML forecasting models Of the three most recent ML-based forecasting models in the past year, Keisler (2022)’s model, FCN (Kurth et al., 2022; Pathak et al., 2022), and Pangu-Weather (Bi et al.”
FuXi: a cascade machine learning forecasting system for 15-day global weather forecast (2023, methodology): “As a remarkable breakthrough, the FourCastNet model [12] is the first of its kind to provide high-resolution global weather forecasts of 0.”
Fourier Neural Operator with Learned Deformations for PDEs on General Geometries (2022, background): “FNO and its variants are able to simulate many PDEs such as the Navier-Stokes equation and seismic waves, do high-resolution weather forecasts, and predict CO2 migration with unprecedented cost-accuracy tradeoff [13–15].”