Scott Lindstrom wrote:

MOS is also produced for the ETA model.

MOS can be a powerful tool for cases that happen again and
again. For rare events, however, that fall outside the realm
of its statistical database, it can go significantly awry.
In addition, I seem to recall it becomes progressively more
heavily weighted towards climatology as you go out into
the future. MOS works well in benign conditions only to
the extent that benign conditions occurred frequently
when the statistical model was constructed.

MOS statistics are continually updated for both the GFS and Eta models.
NGM MOS is frozen. When Eta MOS first appeared a few years ago, it
did a poor job. Now, it often does better then the GFS and NGM.

-db-

+-------------------------------------------------------------+
| David O. Blanchard Flagstaff, Arizona |
+-------------------------------------------------------------+