NZ Weather Forum

Mine is Twiggy?

No, seriously, I like ECMWF because it can be fairly accurate in the long run.

JohnGaul
NZThS

Mainly use my gut instinct.. failing that, GFS and EC ive found are the most reliable.. GASP and NOGAPS are absolutely (swear word) hopeless specially on the long date model runs.. tweak with the variables and see what they throw at you.

ED: Should add that ive never trusted a combination of any above and there are so many met models out there to play with its not funny.. but i really cannot fail the first 2 mentioned.

GFS all the way because it matches MS 7 day maps and this is nearly always right!

ECMWF is proven to be the best. Follow that with UKMO. GFS' poor resolution can't handle the terrain of NZ hence must always be the worst.

Nothing does NZ well. We can only use these for a general picture. I find GFS websites easier to use and higher resolution (looks cooler haha) and it does well at picking up on the main idea of whats going to happen. Its not until something wild pops up on GFS that ill look at others and even then i only look at them generally. We are far to small for a computer model to be accurate.

GFS' poor resolution can't handle the terrain of NZ hence must always be the worst.

Don't know about that. It picks up surface based instability over land quite well in the summer afternoons. I could go on with examples, it does okay. But like any model you have to treat it as model output rather than a forecast.

NZstorm wrote:

GFS' poor resolution can't handle the terrain of NZ hence must always be the worst.

Don't know about that. It picks up surface based instability over land quite well in the summer afternoons. I could go on with examples, it does okay. But like any model you have to treat it as model output rather than a forecast.

GEM is the worlds best for surface based instability.Quite a few applications are now either based on or use the GEM model.

I am an unskilled weather model surfer that finds the interplay between weather systems really interesting. I dont favour one model over another as I find that some models cope well with some systems and not others. Eg NOGAPS was fairly close to the mark on the low pressure system about the affect the NI as early as a week ago and GFS was way off yet it could easily be the other way around for the next system at the end of next week. I suppose what i do is surf the models and if there is general agreement in a particular scenario then there seems to be a higher degree of predictability. When there is divergence in the models like there was earlier this week the predictability goes out the window in my book.

There was mention of the GEM model above. I havent been able to find one for NZ. If one exists I would be grateful if someone could post a link.

Each model is different and deserves a look. Though sometimes I don't bother looking at the BoM's GASP which is pretty much redundant these days. This year, the ECMWF model hasn't been peforming that well compared to other years... but then again the GFS hasn't feared well either.

It's important to look at all models when really looking into an event so you yourself can make a consensus on the output trends and then consider which output is junk and what isn't.

In regard to the "poor resolution" claim of the GFS... well, here's where the use of an MM5 setup works well to use the GFS' boundary conditions and then run it through a nested domain with higher resolution. I'm pretty sure MetService do this with the UKMO with their own 'CHAMPS' MM5 model.

hokimalone wrote:There was mention of the GEM model above. I havent been able to find one for NZ. If one exists I would be grateful if someone could post a link.

Here: http://expert.weatheronline.net.nz/date ... /pslv.html

Foggy wrote: In regard to the "poor resolution" claim of the GFS... well, here's where the use of an MM5 setup works well to use the GFS' boundary conditions and then run it through a nested domain with higher resolution. I'm pretty sure MetService do this with the UKMO with their own 'CHAMPS' MM5 model.

hokimalone wrote:There was mention of the GEM model above. I havent been able to find one for NZ. If one exists I would be grateful if someone could post a link.

Here: http://expert.weatheronline.net.nz/date ... /pslv.html

I find meso scale models pretty interesting, but they can also go wildly wrong and be worse than the global scale models.
The other thing is that MM5 is a dead-end now, really WRF is where it's at.

thanks for the link Foggy. Another model to peruse.