INTELLIGENT USE OF THE AVN/MRF
Why models have forecast problems
The way the physics are approximated can lead to model errors, for example
A NUMBER OF AVN/MRF PERFORMANCE CHARACTERISTICS HAVE CHANGED IN THE PAST YEAR.
AVN/MRF Still Often Has Problems Handling Upslope Events
About 75% of the AVN Rainfall Over the OK Panhandle Was Grid-scale Precipitation (Not Convection).
Another Case: AVN Wraps Low Too Far North And West. Both Surface and 500 mb Lows Are Too Deep.
Aviation Model handling of 500 mb trough
BIAS COMPARISON OF 12-36 HR MRF AND EARLY ETA FORECASTS
The MRF and MRFX spin-up precipitation bombs and tropical systems erroneously at all time ranges.
MRF RELATIVE
HUMIDITY
(pressure -Vs- time)
BEFORE 7/21 AFTER 7/21
MRF THETA-E
(pressure -Vs- time)
BEFORE
7/21 AFTER 7/21
During Oct-Feb 1999, the AVN/MRF now had a bias of
During OCT-JAN 1999 4 INCH OR GREATER ISOHYETS
SINCE THE CHANGES IN JUNE AND JULY, THE ETA BIAS FOR HEAVY AMOUNTS HAS DECREASED SIGNIFICANTLY
One of the typical errors of the AVN is to predict convective systems too far north
The AVN/MRF may be right on the synoptic scale features but cannot handle outflow boundaries, etc.
Despite these problems, the AVN model was the best model during the winter
LOWS TO THE LEE OF THE ROCKIES
THE NGM AND AVN/MRF HAVE SERIOUS PROBLEMS WITH ARCTIC AIRMASSES
Why models have problems with arctic airmasses
MRF PERFORMANCE FOR 3-5 DAY FORECASTS
MRF PERFORMACE FOR 3-5 DAY FORECASTS (CONT)
A VERIFICATION OF THE 500 H FORECASTS FOR DAY 4
You need to know the characteristics of the MRF MOS guidance. Stations included in MOS
MOS verification for Northwest (left) and northern Plains (right)
Verification of MOS POPS for Great Lakes Region (top) and Northeast (bottom)
NUMBER OF DAYS WHEN THE TEMPERATURE ERROR WAS 8 DEGREES OR GREATER ON DAY 4
THE PERCENT OF THE NUMBER OF DAYS (JAN-MAR) HPC FORECASTERS REDUCE THE ERROR BELOW 8 DEGREES F