Model runs

[corpusbreeze]

I was wondering when the next update on the models will come out. Have never understood when to get the next run.Also am looking forward to the new mm5 run since it was the most interesting of all of them. Never was a model fan really, but seems to be the only thing to look forward to since the Caribbean wave cant gets act together. I'm sorry ...just bored.

[corpusbreeze]

Well the mm5 is out looks like another Florida storm if this model is any good.

[Derecho]

Well the mm5 is out looks like another Florida storm if this model is any good.

I wouldn't take it particularly seriously, ESPECIALLY for cyclogenesis (tropical formation)...it seems to spin up boguscanes like crazy, similar to the old MRF.

[Steve]

Here's a fairly good link to the WRF. I haven't followed it much in the past until I switched from PSU to FSU for the model runs. Just click on a photo. I also don't know what the reliability is or if they run it based on another model. But it's still a cool link to have.

http://www.ghcc.msfc.nasa.gov/Model/model_mm5.html

Steve

[Wthrman13]

One thing to keep in mind, and what can get confusing, is that a lot of these community and research models, such as MM5 and WRF, are run by many different groups/agencies using different configurations of the model. In fact, the WRF model in particular is designed with several different "swappable" configuration options, such as the option to choose between two completely different frameworks for solving the basic dynamical equations, essentially making it two (or more) different models in one. The WRF model output that you get from UCAR is not the same as you get from that NASA link, for example. In contrast, whenever you look at GFS, NAM, or RUC output, you can be (almost) assured that you are seeing output from the same run of the same model.

This also spills over into evaluating the model output. Saying, "The WRF sucks for tropical cyclones" may be true of one configuration of the model, but not for another. It really depends. Personally, at this point I put much more stock in the operational model forecasts, and not so much in the experimental runs of these other models, for the very reason that they are experimental. I still look at them though, because they are interesting and often represent the leading edge of research into high-resolution modeling. Just don't expect too much out of them... yet.

[Hurricaneman]

The MM5 I would discount as of now

[dhweather]

One thing to keep in mind, and what can get confusing, is that a lot of these community and research models, such as MM5 and WRF, are run by many different groups/agencies using different configurations of the model. In fact, the WRF model in particular is designed with several different "swappable" configuration options, such as the option to choose between two completely different frameworks for solving the basic dynamical equations, essentially making it two (or more) different models in one. The WRF model output that you get from UCAR is not the same as you get from that NASA link, for example. In contrast, whenever you look at GFS, NAM, or RUC output, you can be (almost) assured that you are seeing output from the same run of the same model.

This also spills over into evaluating the model output. Saying, "The WRF sucks for tropical cyclones" may be true of one configuration of the model, but not for another. It really depends. Personally, at this point I put much more stock in the operational model forecasts, and not so much in the experimental runs of these other models, for the very reason that they are experimental. I still look at them though, because they are interesting and often represent the leading edge of research into high-resolution modeling. Just don't expect too much out of them... yet.

Derek is dead on here - a lot of agencies run the same models,
with configuration "tweaks" or parameters for the mission of
the agency or its customer.