The data and methods used to produce the reanalyses and reforecasts in newest version of NOAA’s Global Ensemble Forecast System

Illustration of a server room corridor and programming nodes

Reanalyses and reforecasts for the most recent version of NOAA’s Global Ensemble Forecast System, version 12 (GEFSv12), are now available to NOAA scientists and external customers. The journal Monthly Weather Review recently published two articles co-authored by NOAA and Cooperative Institute researchers (including the Physical Sciences Lab and CIRES), which describe the data and methods that were used to produce the latest reanalyses and reforecasts, and how to access them. These data are more accurate and reliable than analyses and model forecast guidance from previous-generation systems, and may also facilitate other scientific inquiries such as the performance of the current operational system during past significant weather events.

The reanalysis provided 20 years (2000-2019) of initial conditions for starting retrospective ensemble forecasts (reforecasts) of GEFSv12. Through comparison of the reforecasts with observations, it is possible to detect systematic errors and correct them in real-time guidance, dramatically improving their skill and reliability. Reforecast data are available at Amazon Web Services for free under NOAA’s Big Data Program. The reforecasts are being used inside and outside NOAA for applications such as improving the Climate Prediction Center’s 6-10 day and 8-14 day forecasts, for improving hydrologic predictions, and many other applications.

The public dissemination of reanalysis and reforecast data for NOAA’s Global Ensemble Forecast System version 12 demonstrates the organizational commitment to provide data for the development of value-added weather forecast applications inside and outside NOAA.

Hamill, T. M. (PSL), et al. (January 2022): The reanalysis for the Global Ensemble Forecast System, version 12. Mon. Wea. Rev.,
Guan, H. (NCEP/EMC), et al. (March 2022): GEFSv12 reforecast dataset for supporting subseasonal and hydrometeorological applications. Mon. Wea. Rev.,