A daily climatology from 1948-2007 was computed using NCAR/NCEP Reanalysis data. A 15-day centered average is then used for each day. An average 500 hPa height and a standard deviation of height for each day was computed, as well as an average 850 hPa temperature and standard deviation of temperature. Other fields include winds, precipitable water and moisture flux.
Standard deviation or s is computed by the following formula..
s =square root of (the average of heights2 - (average height)2 )
The standard deviations or s values for each day shows a latitudinal and seasonal varience with the greatest standard deviations found over higher latitudes, particularly over the Northern Pacific and the Aleutians as well as the western North Atlantic. It is important to be aware of these variences not only because they typically are frequently aligned with ensemble forecast uncertainties, but they also give the forecaster an indication of how many meters (degrees) above or below average a height (temperature) field must be to produce a significant number of standard deviations from climatology.
The number of standard deviations from climatology is computed by the following formula..
# of standard deviations = (forecast height - average height) ÷s
This calculation is perfomed with the forecast
model data from the GFS, ECMWF, UKMET, CMC, NAM, SREFMEANS, ECENS and CMCE.