Climate Forecast System (CFS)

Go to:


What is the CFS?

Using the CFS

Purpose of learning about the CFS

On August 24, 2004, the National Oceanographic and Atmospheric Administration (NOAA) introduced the new Climate Forecast System (CFS), as part of a press release on the prediction of a new El Niño in the equatorial Pacific. The CFS, developed by the Environmental Modeling Center (EMC) of the National Centers for Environmental Prediction (NCEP), provides a state-of-the-art dynamical model basis for predicting evolution of the atmosphere-ocean-land system on a time scale of months and lays a foundation for future improvements in monthly and seasonal forecasts. The purpose of this introduction is to provide information to the National Weather Service (NWS) Weather Forecast Office (WFO) staff, particularly the WFO climate focal point, so that they can field inquiries about the CFS from NWS customers.

After viewing this introductory training, the student should be able to:

Links are provided to other resources on the CFS and related topics throughout this web document.  There is also a reference list giving additional resources below. The NWP button at the top of every page takes you to the COMET NWP Operational Model Matrix, and the Climate button takes you to the Climate Services Professional Development Series. The sequence through this training follows the links at the top of the page.

Bridging from weather forecasting to climate forecasting

The CFS uses the same basic atmospheric model as the current National Centers for Environmental Prediction (NCEP) Global Forecast System model.  It is run at coarser resolution so that its forecasts are available in a timely manner, but the model numerical methods and physical processes are very similar.  Using the same atmospheric model in the CFS as in shorter-range weather forecasts provides a “bridge” between weather and climate modeling. In both weather and climate forecasting, a model solution is just one of many pieces of information considered by the forecaster, and the model solution must be used intelligently, taking into account the model limitations. One key difference is that the climate forecaster is interested in the overall essence of the weather over weeks or months rather than in daily or hourly details that the model cannot possibly predict accurately months in advance.

New techniques in the CFS

In order to extend an NWP model solution to several months and obtain reasonable climate statistics, three additional steps are necessary:

  1. Ocean-atmosphere coupling — allows the sea surface temperatures (SSTs) to evolve based on ocean dynamics which include the wind stress from the atmospheric model solution. These SSTs feed back on the atmosphere, such as by influencing the location of tropical convection. Without this coupling, the climate model would not be able to predict the evolution of the El Nino - Southern Oscillation (ENSO) cycle, which profoundly impacts the predominant global weather patterns over a month or season. Ocean-atmosphere coupling is also used in the GFDL hurricane model run by NCEP.
  2. Using ensembles — multiple forecasts must be used together as an ensemble, because the specific weather patterns predicted by any one model run months in advance will not be reliable. Combining many forecasts allows the probability of different scenarios to be estimated or a consistent signal to be detected. Ensembles add value to weather forecasts in the shorter range too, as discussed in the COMET ensemble module.
  3. Correcting predictions based on retrospective forecasts — model biases become even more pronounced at long time ranges than in the shorter weather forecasts, so the forecasts need to be corrected. The CFS corrections are determined by verifying a large sample of model forecasts, obtained by running decades of retrospective cases, a procedure sometimes also called reforecasting.

Where to go for more information

The official CFS website is at :

Questions about the CFS should be directed to

The following resources provide more information on the CFS. References marked by a * are entirely about the CFS. Others provide related information or context and are offered because there are not yet many publications about the CFS. More will be added over time. The first list contains research publications, while the second list contains somewhat simpler presentations also aimed at a professional scientific audience. The third list, however, is geared toward the general public, though a scientific background would help in understanding the material.

Professional journals and publications

Behringer, D. W., M. Ji, and A. Leetmaa, 1998: An improved coupled model for ENSO prediction and implications for ocean initialization. Part I: The ocean data assimilation system. Mon. Wea. Rev., 126, 1013-1021

Hamill, T. M., J. S. Whitaker, and X. Wei, 2004: Ensemble reforecasting: Improving medium-range forecast skill using retrospective forecasts. Mon. Wea. Rev., 132, 1434-1447.

Kanamitsu, M., W. Ebisuzaki, J. Woollen, S-K. Yang, J. J. Hnilo, M. Fiorino, and G. L. Potter, 2002: NCEP–DOE AMIP-II Reanalysis (R-2) , Bull. Amer. Meteor. Soc., 83, 1631-1643.

Phelps, M. W., A. Kumar, and J. J. O’Brien, 2004:  Potential predictability in the NCEP CPC dynamical seasonal forecast system. J. Climate, 17, 3775-3785.

S. Saha, S. Nadiga, C. Thiaw, J. Wang, W. Wang, Q. Zhang, H. M. van den Dool, H.-L. Pan, S. Moorthi, D. Behringer, D. Stokes, M. Pena, S. Lord, G. White, W. Ebisuzaki, P. Peng, P. Xie , 2006 : The NCEP Climate Forecast System. Journal of Climate, Vol. 19, No. 15, 3483-3517.

Wang, W., S. Saha, H.-L. Pan, S. Nadiga, and G. White, 2005: Simulation of ENSO in the new NCEP Coupled Forecast System Model. Mon. Wea. Rev, 133, 1574-1593.

References on the web (CPC, CDC, Climate PDS, PowerPoints)

Cooperative Program for Meteorological Education and Training (COMET), cited 2004:  Climate Professional Development Series (PDS), Professional Competency Unit (PCU) 3:  Demonstrate understanding of the basis and methodologies of CPC products. [Available online at ]

National Centers for Environmental Prediction Climate Prediction Center, cited 2004:  Seasonal outlook.  [Available online at ]

O’Lenic, Edward, cited 2004:  Climate Prediction Center review of the NCEP production suite: Changes, performance and plans.  [Available online at ]

*Pan, H.-L., cited 2004:  The NCEP operational Climate Forecast System : configuration, products, and plan for the future.  [Available online at ]

*Saha, S., cited 2004: Documentation of NCEP CFS data files. [Available online at ]

*Thiaw, C. and S. Saha, cited 2005: CFS retrospective Forecast Time Series of Monthly Means in the EMC/NCEP NOMAD public server. [Available online at ]

References for non-meteorologists

*NOAA News Online, cited 2004:  NOAA’s New Global Climate Forecast System Increases confidence that weak El Niño conditions are developing.  [Available online at ]

NOAA Climate Diagnostics Center, Climate Research Spotlight Article: Improving week-2 weather forecasts through "re-forecasting". [Available online at ] (discusses calibration with retrospective forecasts but for 2-week GFS forecasts without ocean coupling)


Dr. William Bua, UCAR/COMET
Dr. Stephen Jascourt, UCAR/COMET

With assistance from
Dr. Suranjana Saha, NCEP/EMC
Dr. Marina Timofeyeva, NWS/OCWWS
Dr. Huug Vandendool, NCEP/CPC

Thanks also to Jeff Waldstreicher, NWS/ER/SSD for motivating this training and Heather Hauser, NWS/ER/SSD for reviewing it