The National Centers for Environmental Prediction (NCEP) has launched the Conventional Observation Reanalysis (CORe), a new operational dataset designed to support monitoring of the Earth system. CORe was developed through collaboration among the Climate Prediction Center (CPC), the Environmental Modeling Center, and the Physical Sciences Laboratory, with support from the Weather Program Office’s Subseasonal-to-Seasonal Program. Implemented on March 19, 2026, CORe provides atmospheric analyses every three hours from 1950 to the present, with updates available after a one-day delay. These analyses are publicly available on the cloud through the
CORe is built on the GFSv15 modeling framework with an Ensemble Kalman Filter data assimilation system. It is replacing the Climate Data Assimilation System (CDAS) at CPC for both real-time monitoring and retrospective applications, including forecast verification and the development of statistical and artificial intelligence tools. The dataset, spanning historical conditions through near real time, is publicly available on the cloud through the NOAA Open Data Dissemination program.
One of the deficiencies of reanalysis datasets is the accurate representation of precipitation patterns. The figure below shows that CORe reproduces the intraseasonal precipitation variability well compared with observations and substantially outperforms CDAS, reflecting improved model physics, data assimilation, and large-scale circulation.
CPC has begun incorporating CORe into its operational and monitoring products, replacing CDAS across its workflows. This transition marks an important step in NOAA’s efforts to modernize its analysis capabilities while maintaining consistency with long-term records and improving the foundation for forecast evaluation and development. Additional evaluation results and documentation are available through the Climate Prediction Center website.