Dennis M. Rodgers wrote:
Edgar D. wrote:

Hello.

Can someone tell me the advantages/disadvantages of the following
numerical
weather models in terms of weather forecasting?

ETA
MRF
MOS
NOGAPS
GEM

Specifically, I would like to know what the strengths and weaknesses of
these models are when using them to produce a weather forecast. If I am
missing any models, feel free to add them.

Thanks

Ed



[snip]
MOS is not a model. It stands for Model Output Statistics. MOS is
forecast guidance produced from the NGM and GFS models. It is produced
with multivariate linear regression of model output parameters and
cranks out temp and precip forecasts. Works well in benign conditions.
Works less well around weather systems.


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.

Scott