Climate Prediction Center - seasonal Outlook
Skip Navigation Links www.nws.noaa.gov 
NOAA logo - Click to go to the NOAA home page National Weather Service   NWS logo - Click to go to the NWS home page
Climate Prediction Center
 
 

CPC Search
Model Guidance Tools

Probabilistic Extremes Tool

Hazard Assessments

U.S. Hazards Outlook

Global Tropics Hazards and Benefits Outlook

 
HOME > Outlook Maps> Hazard Assessment> Week-2 Extremes Tool> Extremes Tool About
 
Week-2 Global Probabilistic Extremes Forecast Tool Description

 

About the tool

The 8-14 day (further referred to as week-2) Probabilistic Extremes GEFS Reforecast Tool is a model guidance tool that applies statistical adjustments to raw Global Ensemble Forecast System (GEFS) model forecasts. These statistical adjustments typically provide more skillful forecasts during the week-2 period, compared to raw model output. Model data is both bias-corrected (mean error is removed) and calibrated (probability adjusted) using the long-term GEFS reforecast data. The ensemble regression method (Unger, 2009) is used to calibrate the probabilities. Bias correction and calibration are deemed as necessary post-processing steps to producing more skillful forecasts, especially beyond the week-1 time frame.

The Week-2 Probabilistic Extremes GEFS Reforecast Tool is the primary guidance used by the week-2 U.S. Hazards forecasters at the Climate Prediction Center (CPC). This tool, in addition to other tools and model guidance, helps to inform CPC's official forecast maps. The back-end data is produced at the CPC and displayed using the ArcGIS web server hosted by the NOAA GeoPlatform (NOAA's instance of ArcGIS Online).

The Week-2 Probabilistic Extremes GEFS Reforecast Tool consists of 4 variables:

(1) Maximum Temperature (tmax) - Daily forecasts during week-2 of maximum temperatures exceeding the 85th percentile of the 30-year historical distribution, 80°F, 90°F, 100°F, and 110°F. Valid bounding times are 06Z to 06Z for each day.

(2) Minimum Temperature (tmin) - Daily forecasts during week-2 of minimum temperatures falling below the 15th percentile of the 30-year historical distribution, -40°F, 0°F, 28°F, 32°F, and 40°F, and exceeding 80°F and 85°F. Valid bounding times are 06Z to 06Z for each day.

(3) Accumulated Precipitation (precip) - Forecasts of accumulated precipitation during the 8-10, 10-12, and 12-14 day periods exceeding the 85th percentile of the historical distribution, 1", 2", and 4". Valid bounding times are 12Z to 12Z for the 3-day periods.

(4) Maximum Wind (wmax) - Forecasts of maximum wind speed at any time during the 8-10, 10-12, and 12-14 day periods exceeding the 85th percentile of the historical distribution, 25 mph, 40 mph, and 50 mph. Contours represent the forecast probability of the maximum instantaneous wind speed over the given 3-day period exceeding the percentile or threshold for a given location. Valid bounding times are 12Z to 12Z for the 3-day periods.

All variable forecasts are based on the 00Z update cycle of the GEFS of the current day.


Interpretation

The contours represent forecast probabilities reaching values that are greater than or less than various percentiles or thresholds at given locations, depending on the variable for the valid day selected. Contours in bold represent forecast probabilities typically used as risk level thresholds in the Week-2 U.S. hazards outlook (https://www.cpc.ncep.noaa.gov/products/predictions/threats/threats.php).
The percentiles are calculated based on a 30-year normal/climatology (i.e. historical mean) from 1991 to 2020. Percentiles were selected based on what would likely produce sufficiently skillful forecasts in the extreme range, indicating what may be considered hazardous.
Threshold values (e.g. actual temperatures, precipitation amounts, or wind speeds) were chosen based on conditions that may be deemed hazardous. Each variable’s forecast time aggregation was selected based on those that likely have the most predictability at the week-2 lead associated with model skill. Forecasts are provided across the global domain.


References
Unger, D.A., H, H. van den Dool, E. O’Lenic, and D. Collins, 2009: Ensemble Regression. Mon. Wea. Rev., 137, 2365–2379, doi: 10.1175/2008MWR2605.1.



Feedback/questions send to: Melissa Ou

NOAA/ National Weather Service
National Centers for Environmental Prediction
Climate Prediction Center
5830 University Research Court
College Park, Maryland 20740
Page Author: Climate Prediction Center Internet Team
Page last modified: December 12, 2005
Disclaimer
Information Quality
Credits
Glossary
Privacy Policy
Freedom of Information Act (FOIA)
About Us
Career Opportunities