NOAA’s National Centers for Environmental Information (NCEI), National Centers for Environmental Prediction (NCEP), and the Climate Program Office are jointly announcing the completion of the public archive of the reforecasts and reanalysis from the Climate Forecast System (CFSv2)—the current NCEP operational system for monthly and seasonal forecasts. Scientists use forecast models like CFSv2 to “reanalyze” archived observations and forecast past conditions, which allows them to create datasets describing global interactions between Earth’s oceans, land, and atmosphere.
Accurate knowledge of the historical condition of Earth’s climate is critical for a number of applications. However, the observing systems that monitor various Earth system indicators are sparse, both in placement and the length and consistency of data records. Reanalysis uses climate models to fill in the spatial and temporal gaps between observations by using the models’ physically consistent representation of the Earth system. A reanalysis dataset thus provides the most consistent view of past Earth system conditions. Reforecasts are related to reanalysis data, except they are focused on improving current climate forecasts. By going back and making forecasts for historical time periods, scientists can evaluate the performance of a particular forecast system and calibrate it for current forecasts.
These reanalysis data and reforecasts are critically important to the research and forecasting community. The research community uses reanalysis to understand climate variability and impacts in a historic context. Dr. Jim Kinter, a user and Director of the Center for Ocean-Land-Atmosphere Studies (COLA), said, “The complete set, extending over more than 30 years of real climate states, is an invaluable resource for both statistical analysis of characteristics of climate forecasts and as a control set for conducting numerical sensitivity experiments.” According to Dr. Kinter, “All of these are critical issues in climate prediction research that could not be addressed adequately without this dataset from NCEI.”
Data available in this archive will enable research that underpins the understanding of climate variability, improve seasonal and monthly predictions of Earth system conditions such as the current El Niño, and support water resource forecasting. Dr. Eric Wood of Princeton University said, “[This dataset] is a critical resource for advancing drought and water resources research, and for experimenting and developing new prediction systems to help the nation cope with seasonal conditions.”
Beyond the research community, the private sector uses reforecast and reanalysis data from CFSv2 extensively, especially in energy, agriculture, water resource, insurance, and reinsurance applications. John Dutton of Prescient Weather states, “Prescient Weather relies on these data to support a wide range of innovative products. CFS reforecasts and reanalyses provide an essential component of our World Climate Service, which predicts both traditional weather variables and industry impact variables on timescales from two weeks to six months into the future.”
Developed jointly by NCEI, NCEP, and CPO, the archive is accessible to the user community through the NOAA National Operational Model Archive and Distribution System (NOMADS). The complete CFSv2 archive, which totals 864 terabytes of data, currently stands as a unique public resource and is the most comprehensive archive of forecast data available internationally from a single forecast system. Over 248 million individual downloads occurred in FY 2015, accounting for 1.325 petabytes of data.
The CFSv2 archive includes NCEP’s operational model data, along with the broader reanalysis and reforecast data. Time periods available include:
- Reanalysis data for 1979–2011
- Reforecasts for 1982–2011
- Operational NCEP CFSv2 model analysis and forecasts covering April 2011–December 2014
The archive will be expanded in the future to include additional operational data up until near-real time. For more information on model data products, check out the NOMADS project website. And, for a list of CFSv2 data products, see our Climate Forecast System web page.