NWC REU 2013
May 22 - July 30



Photo of author

An Evaluation of the Climate Forecast System Version 2 as an Extended Range Forecasting Tool in the Storm Prediction Center

Joshua Crittenden, Harold Brooks, Greg Carbin, Sean Crowell, and Patrick Marsh


What is already known:

  • Over half of the Storm Prediction Center’s (SPC) Day 4–8 Severe Weather Outlooks for the CONUS are a forecast of “Predictability Too Low.”
  • The Climate Forecast System Version 2 (CFSv2) is a fully coupled ocean-land-atmosphere seasonal prediction model from which severe weather proxies can be derived.
  • This study investigates whether consistency of severe weather proxies in CFSv2 forecasts can aid SPC forecaster decision making for Day 4–8 outlooks.

What this study adds:

  • Statistics of SPC Day 4–8 Severe Weather Outlooks were calculated for January – June 2013; the focus being on May and June which contained contrasting cases of severe weather forecasting in the SPC.
  • Severe weather forecasts were approximated by calculating the Supercell Composite Parameter from CFSv2 output.
  • SPC Outlooks and Filtered Storm Reports were used to assess forecast quality from the CFSv2 for May and June 2013.
  • Cases studied indicate consistency of a severe weather proxy in CFSv2 output may assist SPC forecasters in providing more specific severe weather information in Day 4–5 forecasts.


As of today, extended range forecasts cannot be made on a consistent day to day basis. The ability of forecasters to predict severe weather beyond a three day lead time is limited. If it is made possible for forecasters to make reliable and consistent extended range forecasts, then the safety of the public will be enhanced by severe weather warnings several days in advance.


In order to potentially give forecasters a new tool in assisting with extended range forecasting of severe weather, the Climate Forecast System Version 2 (CFSv2) and its forecasts are being examined and compared with the Storm Prediction Center (SPC) Day 4–8 forecasts and also compared with actual reported events.


Granted that there are days without severe weather, few of SPC Day 4–8 Severe Weather Outlooks have actual forecasts. The CFSv2 has shown an ability to reliably forecast severe (or lack of severe) weather with a day four lead time and moderate reliability at day five. Although the CFSv2’s capability to forecast reliably beyond day five is, to some degree, limited, in this paper it is shown that the CFSv2 does have potential as an extended range forecasting tool.

Full Paper [PDF]