NMM UK Model now higher res and out to 54 hours

[Paul]

Hi all, we've been working on this for some time and now it's online and ready to be used.

The new Netweather Hi-res (0.09 degree / 12km model) is now available to use within NW Extra, this new mesoscale model offers significantly more detail than the GFS charts (which run at 0.5 degree / ~60km), is run twice a day out to 48 hours and has 1 hour timeslots.

The improved resolution and dynamics of this model, mean increased accuracy and detail when forecasting all weather situations, particularly localised rain events, convective situations, rain/sleet snow setups and so on.

Currently the viewer is in its beta phase while we fine tune things and add new charts on a regular basis, and is available to our full extra subscribers. You can view it here:

http://www.netweather.tv/secure/cgi-bin/pr...ction=nmm;sess=

Here are a few screenshots from the model:

[Ross B]

Nice one Paul, looks very useful :lol:

I also like the 3rd chart you posted there in particular :lol:

[Cookie]

looks good guys :lol:

[Stuart]

Look Good paul :lol:

[Zenarcher]

Looks great :lol:

[James M]

That looks brilliant Paul, would have to save up if was to get it though.

Edited July 27, 2008 by James M

[shuggee]

Brilliant. And here's the but....

The T850 + SLP 1 & 4mb maps seem to be identical Paul?

[Paul]

Ah, that'd be because I put identical file names in the menu, try now!

[shuggee]

Come on, admit it, you did that on purpose and you and Karl have a bet on on how long it would take somebody to notice?

Do I win a prize?

[Paul]

Yep, Karl said he'd give it to you at the team meet up.

[shuggee]

Fnar fnar - that will be interesting

[J07]

Hi all, we've been working on this for some time and now it's online and ready to be used.

The new Netweather Hi-res (0.09 degree / 12km model) is now available to use within NW Extra, this new mesoscale model offers significantly more detail than the GFS charts (which run at 0.5 degree / ~60km), is run twice a day out to 48 hours and has 1 hour timeslots.

The improved resolution and dynamics of this model, mean increased accuracy and detail when forecasting all weather situations, particularly localised rain events, convective situations, rain/sleet snow setups and so on.

This is good stuff, I hear that WRF is generally quite an improvement over MM5. However, I think the last paragraph should be read with caution. These mesoscale models can fail, they do not necessarily always give improved accuracy. Sometimes a more coarse grid spacing can actually give better results. It all depends on the subtle nuances of the mesoscale model itself. eg the resolution of the input data (such as sea surface temperatures), but also what the convective parametrisation scheme is being used within the model (I assume there is one).

Remember also that even with 12km grid spacing, it will not capture the terrain of Britain particularly well. It will smooth out a lot of features. Given that the UK is mostly flat with the odd undulation, this is not too much of a concern for many- but in Scotland it could be troublesome.

Were there a 3km or 6km model of course that would be better in theory....but again, I hear there are difficulties with these and that sometimes they give rubbish outputs compared with lower resolutions.

[VillagePlank]

It depends on what is being measured, I guess. If I had a 1000 miles model and said it's gonna rain today in the GB part of the grid, I'd be right 99.9999999 (rec) of the time, so an extremely low-res model would have a greater hit rate than one, say, 500 miles, and, of course, as you 'zoom' in the less, by chance, you've got of getting it right.

Is it right to presume that the reason it only goes out to 48 hours is because of the extent of the horizontal sampling around the UK (presume you sample from Eastern Seaboard to Russia, and North Pole to Equator?) - or is it a secret? :lol:

Great idea, though, might well be signing up the NetWE in the near future!! :lol:

Edited August 13, 2008 by VillagePlank

[Paul]

Hi VP, the 48 hour running time is literally because of the run-time, we'll probably look to increase that at some point as we upgrade hardware etc.

We've also made an improvement to the model overnight last night, we've increased the resolution from 12km to 8km which gives a fair amount more detail.

[multi cellular thunderstorm]

Hi VP, the 48 hour running time is literally because of the run-time, we'll probably look to increase that at some point as we upgrade hardware etc.

We've also made an improvement to the model overnight last night, we've increased the resolution from 12km to 8km which gives a fair amount more detail.

I noticed the change, especially on the wind charts, it should be much better for picking out convergence lines etc. Before the update it didn't seem to have many arrows, a bit like the GFS after +180.

[Paul]

Evening all, just to let you know we have now added some extra hours to the nmm forecasts, taking it out to 54 hours (in 1 hour time slots) and at 8km resolution

View it here:

http://www.netweather.tv/secure/cgi-bin/pr...ction=nmm;sess=

[Paul]

For those with a lite subscription, we've added a lite version of the nmm viewer to your subscription, so you can access key charts from the model from here on in

For those who already subscribe, you can find the NMM Lite viewer here:

http://www.netweather.tv/secure/cgi-bin/pr...n=nmmlite;sess=

For those who would like more information about the subscription options, click here:

https://www.netweather.tv/secure/cgi-bin/login.pl