GEOS (Goddard Earth Observing System)

GEOS (Goddard Earth Observing System)
Name	GEOS (Goddard Earth Observing System)
Developer	NASA
Released	1990s
Programming language	Fortran, C, Python
Operating system	Unix-like
License	Proprietary (NASA)

Contents

GEOS (Goddard Earth Observing System) is a coupled atmosphere–ocean–land–chemistry modeling and data assimilation system developed and maintained by NASA's Goddard Space Flight Center and collaborators for weather prediction, climate research, and Earth system science. The system integrates satellite observations from programs such as Landsat, MODIS, and Aqua with physical models used by agencies including the National Oceanic and Atmospheric Administration, the European Space Agency, and academic centers like Massachusetts Institute of Technology and Georgetown University. GEOS supports operational products, reanalyses, and specialized research efforts for institutions such as NOAA/NCEP and the Jet Propulsion Laboratory.

Overview

GEOS couples dynamic cores, physical parameterizations, and chemical modules to simulate the Atmosphere of Earth, Ocean, Cryosphere, and Biosphere across time scales from hours to centuries. It ingests remote sensing streams from platforms including Terra (satellite), Aqua (satellite), Suomi NPP, and Jason-3 while interfacing with in situ networks such as Argo (oceanography), Global Precipitation Measurement, and FLUXNET. The system produces reanalysis datasets comparable to efforts like ERA-Interim and MERRA and interoperates with standards from World Meteorological Organization and Group on Earth Observations initiatives.

Development began at Goddard Space Flight Center in the 1990s building on numerical weather prediction advances from groups including European Centre for Medium-Range Weather Forecasts and National Centers for Environmental Prediction. Early GEOS versions incorporated assimilation techniques influenced by work at University of Maryland and Scripps Institution of Oceanography, and satellite data streams from missions such as TOPEX/Poseidon and TRMM. Collaborations expanded to incorporate chemistry modules developed in partnership with Harvard University and California Institute of Technology, and ocean coupling with efforts at Woods Hole Oceanographic Institution. Over successive generations GEOS integrated ensemble methods pioneered at NCAR and variational/ensemble hybrid schemes influenced by Naval Research Laboratory research.

GEOS architecture consists of a dynamical core, physical parameterization suite, chemical transport models, land surface model, and ocean/sea ice components. The dynamical core variants share heritage with models at Princeton University and University of Reading; physics suites include convection and microphysics schemes similar to those used by UK Met Office and Canadian Meteorological Centre. Chemistry components implement chemical mechanisms developed by groups at NOAA Chemical Sciences Laboratory and Max Planck Institute for Chemistry. The land model uses parameterizations comparable to Community Land Model approaches and ingests vegetation products from MODIS and soil datasets produced by Global Soil Data Task. Coupling uses frameworks inspired by Earth System Modeling Framework and ESMF conventions to interoperate with external models from European Centre for Medium-Range Weather Forecasts and Met Office Hadley Centre.

GEOS employs advanced data assimilation including three-dimensional variational, four-dimensional variational, and ensemble Kalman filter techniques developed in concert with researchers at University of Washington and University of Colorado Boulder. The assimilation system assimilates radiances from instruments on NOAA polar-orbiting satellites and geostationary platforms like GOES-R Series, retrievals from AIRS, and limb sounders such as OMPS. Chemical data assimilation incorporates observations from Ozone Monitoring Instrument and MLS. Model configurations include atmospheric general circulation, coupled ocean–atmosphere, and chemistry–climate variants, enabling studies comparable to Coupled Model Intercomparison Project experiments and integrated assessments used by the Intergovernmental Panel on Climate Change.

GEOS underpins operational forecasting products, satellite mission support, and retrospective reanalyses used by NOAA, NASA Earth Science Division, and academic researchers at Columbia University and University of California, Berkeley. Applications include air quality forecasting for metropolitan regions such as Los Angeles, volcanic ash tracking for agencies like Federal Aviation Administration, hurricane prediction supporting National Hurricane Center, and carbon-cycle monitoring utilized by CarbonTracker and carbon research at Oak Ridge National Laboratory. GEOS outputs feed downstream systems at European Space Agency and regional modeling centers like Met Éireann and the Bureau of Meteorology.

Performance evaluation uses observational networks from Global Climate Observing System, GCOS, and satellite validation campaigns involving CALIPSO and ICARE. Validation studies compare GEOS reanalyses with datasets such as ERA5, MERRA-2, and paleoclimate reconstructions produced by PAGES investigators. Limitations include sensitivity to satellite data biases identified in studies by Jet Propulsion Laboratory and computational constraints noted by centers like Argonne National Laboratory and Lawrence Livermore National Laboratory, which affect ensemble size and resolution. Representation of boundary-layer processes and regional convection remains an active challenge echoed by research at Massachusetts Institute of Technology and University of Oxford.

Ongoing development emphasizes higher-resolution coupled simulations, improved aerosol–chemistry interactions informed by work at MIT and Pennsylvania State University, and next-generation assimilation leveraging machine learning research from Stanford University and Google DeepMind. Integration with emerging satellite missions such as Sentinel-6 and follow-ons to Aqua and Landsat will expand observational constraints, while collaborations with European Space Agency and NOAA aim to operationalize sub-seasonal to seasonal forecasting capabilities comparable to Copernicus services. Computational scaling efforts involve partnerships with NASA Advanced Supercomputing, Oak Ridge Leadership Computing Facility, and the National Center for Atmospheric Research to enable century-scale ensemble reanalyses and multi-model intercomparison contributions to future Intergovernmental Panel on Climate Change assessments.

Category:Earth system models