NOAA's Global Forecast System

NOAA's Global Forecast System
Name	Global Forecast System
Operator	National Oceanic and Atmospheric Administration
Country	United States
Founded	1980s
Predecessor	Aviation model suite
Type	Numerical weather prediction model
Grid	Global
Resolution	variable
Website	National Weather Service

Contents

NOAA's Global Forecast System is a global numerical weather prediction model operated by the National Oceanic and Atmospheric Administration and run operationally by the National Weather Service and the National Centers for Environmental Prediction. It provides medium-range to extended-range forecasts used by agencies such as the Federal Aviation Administration, United States Air Force, and international partners like the European Centre for Medium-Range Weather Forecasts and the Japan Meteorological Agency. The system ingests observations from satellites such as GOES and MetOp, radiosonde networks like the Global Observing System, and surface networks managed by institutions including the World Meteorological Organization.

Overview

The Global Forecast System is a unified atmospheric model producing deterministic forecasts and ensemble guidance for time horizons from 1 hour to 16 days, supporting operational centers including the National Hurricane Center and Joint Typhoon Warning Center. It couples atmospheric dynamics with parameterizations for physical processes used by services like National Aeronautics and Space Administration in research collaborations and by regional agencies such as the National Oceanography Centre for downstream ocean coupling. Outputs include fields familiar to agencies like the Federal Emergency Management Agency and operators of infrastructure such as Port of New York and New Jersey for planning during events like Hurricane Katrina and Superstorm Sandy.

Development traces to early numerical work at institutions like Princeton University and research programs such as the Global Weather Experiment. Early operational predecessors were influenced by projects at the European Centre for Medium-Range Weather Forecasts and the United Kingdom Met Office. Funding and programmatic oversight have involved agencies including the Office of Naval Research, National Science Foundation, and bilateral collaborations with the Canadian Meteorological Centre. Key milestones relate to the advent of satellite programs like TIROS and NOAA-AVHRR and computing milestones at centers such as the Oak Ridge National Laboratory and National Center for Atmospheric Research.

The GFS is formulated on the primitive equations solved on a global grid, with dynamical cores and physics suites developed in collaboration with laboratories like the Geophysical Fluid Dynamics Laboratory and research groups at Massachusetts Institute of Technology and Colorado State University. Parameterizations address convection studies rooted in work from Louisiana State University, boundary layer schemes influenced by groups at University of Washington, and radiation models reflecting advances at University of Colorado Boulder. The model couples to land-surface modules using datasets maintained by United States Geological Survey and to sea-ice models informed by National Snow and Ice Data Center. For ensemble systems, concepts derive from frameworks developed at Rutgers University and Imperial College London.

GFS data assimilation integrates observations through systems like the Gridpoint Statistical Interpolation and hybrid variational-ensemble methods developed with contributions from the European Centre for Medium-Range Weather Forecasts, Meteorological Service of Canada, and researchers at Scripps Institution of Oceanography. It ingests remote sensing data from satellites including Suomi NPP, NOAA-20, and Sentinel missions, aircraft observations coordinated via IATA procedures, and in-situ measurements from networks such as the Global Drifter Program and International Comprehensive Ocean-Atmosphere Data Set. Assimilation research leverages reanalysis efforts exemplified by the 20th Century Reanalysis Project and joint ventures with institutions like the University of Maryland.

Products include synoptic fields, precipitation forecasts, wind and temperature profiles, and derived indices used by the National Hurricane Center, Hydrometeorological Prediction Center, and international services such as the Australian Bureau of Meteorology and India Meteorological Department. Aviation guidance supports organizations like the International Civil Aviation Organization and Boeing operations. Energy companies, exemplified by ExxonMobil and NextEra Energy, use wind and solar forecasts; agricultural stakeholders from institutions like the Food and Agriculture Organization use seasonal guidance tied to phenomena including El Niño–Southern Oscillation. Emergency management for events such as Typhoon Haiyan and heat waves by authorities like California Department of Water Resources rely on GFS-derived situational awareness.

Verification compares GFS skill against models like the European Centre for Medium-Range Weather Forecasts's model and the Met Office Unified Model, using metrics developed in collaborations with World Meteorological Organization working groups. Performance assessments examine root-mean-square error and anomaly correlation for geopotential height and precipitation against observational networks such as Global Precipitation Measurement and radiosonde arrays managed by NOAA's National Centers for Environmental Information. Long-term studies at institutions including Cornell University and University of Oklahoma document systematic biases and drive improvements implemented after evaluations with partners like the Office of Science and Technology Policy.

Ongoing research involves higher-resolution configurations, stochastic parameterizations inspired by work at École Normale Supérieure, and seamless prediction frameworks promoted by initiatives such as the World Weather Research Programme. Upgrades have included coupling with ocean models developed at NOAA Geophysical Fluid Dynamics Laboratory and assimilation advances influenced by teams at European Space Agency and NASA's Global Modeling and Assimilation Office. Future directions emphasize machine learning techniques explored at Google DeepMind and universities like Stanford University and Massachusetts Institute of Technology, expanded ensemble systems coordinated with Canadian Centre for Climate Modelling and Analysis, and closer integration with global initiatives such as Group on Earth Observations.

Category:Numerical weather prediction models