MetService To Improve Forecasts

[gllitz]

Well, this should hopefully help make them more accurate:

http://www.stuff.co.nz/national/3959661 ... e-accuracy

[melja]

Well, this should hopefully help make them more accurate:

http://www.stuff.co.nz/national/3959661 ... e-accuracy


(Moderators: the above could almost be a separate thread? Will leave it to you to....uhm, moderate....)

Or less accurate, i think they use them to much now.

[Manukau heads obs]

I cant see how it would make things less accurate?
you sure come out with some strange comments melja

[NZstorm]

Any improvement in modeling is a step forward, but I don't think melja is far wrong.

[Richard]

I think the day to day weather forecasting from the MS is very accurate as it is and good on them for wanting to do even better,but its just that they let themselves down badly with the forecasting of regional temperatures,i would like to see them put more effect into improving this area first.

[NZstorm]

Model output shouldn't replace sound forecasting and I think there have been instances in the past where it clearly has. Yes, the temperature forecasts are the best example of this.

[Manukau heads obs]

regional temperature differences can be large at this time of year for overnight temperatures
a higher resolution model should help with that

for example last night in the auckland area, exposed coastal locations recorded overnight temperatures 5C warmer than in land sheltered locations....

[Richard]

Well here's hoping a higher resolution model can help because here in Canterbury it really is poor

[melja]

I cant see how it would make things less accurate?
you sure come out with some strange comments melja

So what, i have differing opinions than the masses but please keep your comments directed at the topic not me as a personal attack.

Anyway what i was meaning was as there relience on the computer to give forcasts grows( It clearly is allready) then the human touch and moderation of these forcasts may get less, it may leed to cases of less accurate forcasts for smaller areas as models are great for a wider picture but here in New zealand with a very narrow land mass and large mountain ranges they clearly need people with a good understanding of local phonmona to put a human touch to the forcasts.

[Manukau heads obs]

So what, i have differing opinions than the masses but please keep your comments directed at the topic not me as a personal attack.

OK
then
how about starting up your own forum where you can post what you like about the M/S so that I do not have to read it here on this forum
because I do not want to come onto this forum and read your posts about the M/S all the time

but I guess my option is to make it so that I do not read your posts anymore

[RWood]

I think temperature forecasting for this country is a nightmare assignment in many situations, probably nowhere more so than in Canterbury. I see a lot of whinging on the matter in this forum, but no scenarios for improvement.

[Dale]

Should really keep out of it but no.. im with RWood on this. This place is a worse than a nightmare to forecast anything proper.. all my years of doing my own back in Aussie, thats a lot easier, bigger land mass, less dynamic scenarios depending on where you are, its not hard. Here, its hit and miss.. like trying to shoot ants with a shotgun.. might blow 20 cartridges but only get 1. Just from my own training alone I apply the rule that whatever Auckland metro is FC overnight per example, i take off 2 or 3 degrees and usually pretty much spot on.. it all boils down to locality & a genuine interest on your own surroundings.. the average Joe looks at the FC or the TV and thinks well this is going to happen but its not always the case.. cant really bash the M/S on this one.

[Simon Culling]

Dont forget that a decrease in grid size doesn't automatically mean the model predictions are going to be any more accurate. The assumptions on which the model is based are more important, i.e. its physical representation of the atmosphere....reducing the grid size will just give you a more detailed output, not necessarily a more accurate one. I would imagine forecasting in New Zealand is not easy due to the topography and the huge expanses of ocean all the way around. The UK at least has the continent of Europe as close as 22 miles to the east !!!

[melja]

So what, i have differing opinions than the masses but please keep your comments directed at the topic not me as a personal attack.

OK
then
how about starting up your own forum where you can post what you like about the M/S so that I do not have to read it here on this forum
because I do not want to come onto this forum and read your posts about the M/S all the time

but I guess my option is to make it so that I do not read your posts anymore

Man got a bee in your bonnet havent you
The post was put in a topic to debate M/S so i cant see a problem and others have said this the place debate these things.
Out off 4 pages i had about 6 posts.

[Michael]

Our metservice doesnt look to bad (as far as free products) compared to this in south africa.Looks quite a wealth of info though if you have the wealth
http://www.weathersa.co.za/web/

[Paul Mallinson]

A model with reduced grid size represents the terrain much better, and therefore the interactions with the land of the model weather will move closer to reality. I agree that picking the right driving (global model) is very important, but once that is done, then the result is a better wind field, that leads to all sorts of outcomes: better sea breeze representations (and hence better temperature forecasts for a place like Christchurch) better rainfall amounts, better hanndling of wind convergence fields that trigger showers and thunderstorms better profiles of the atmosphere in the vertical and on and on.
I think it is much more important to get the significant weather correct than worry too much about whether today's high was off by 2 or 3 degrees. Someone previously said, weather conditions can vary quite a lot across an area. People like farmers and boaties know how to interpret the local area forecast.
When I was a kid growing up in Blenheim, when we heard that a cold southerly with showers or rain was forecast for the Kaikoura coast, we always knew it was likely to turn up in Blenheim as a cloudy cool northeasterly with the odd shower.

Paul

[NZstorm]

from the BOM website

From 17 August 2010, the Bureau of Meteorology will no longer run the GASP, LAPS, TLAPS and Meso-LAPS family of computer forecasting models.

A new computer forecasting model (ACCESS) will be implemented from July 2010 as the replacement. The performance of the ACCESS model represents a significant leap forward in terms of forecasting abilities over the Australian Region.

[Richard]

I think temperature forecasting for this country is a nightmare assignment in many situations, probably nowhere more so than in Canterbury. I see a lot of whinging on the matter in this forum, but no scenarios for improvement.

Its you that has missed the points i've raised before in other threads also,yes i agree that in some situations( i can only think of one),where forecasting temps in Canterbury could be,as you call it a nightmare assignment,that of, if and when the NW wind arrives and how that effects temperature change,i have and never will criticize the MS during such events.
But my main two areas criticism have nothing to do with any nightmare assignment, in fact one of them is when the weather is at its most settled and that is the issue of frost,only twice this winter has Rakaia,Darfield and Waipara been issued a frost forecast,but all those centers would have received at least 30 frosts so far this winter,where's the nightmare in getting that right??
My other point is Waipara's forecasted max temps are,every day, day in, day out, month after month, 2-5 dec cooler than the all the other Canterbury rural centers,in fact Waipara is issued about the same temps as Te Anau and Lumsden.How the bloody hell could they have a established such a successful wine growing industry there if they had a climate as cold as Te Anau and Lumsden. Waipara is NOT colder than those Canterbury rural centers, so how on earth could this also be a nightmare assignment in getting that right Rupert??

[03Stormchaser]

When I was a kid growing up in Blenheim, when we heard that a cold southerly with showers or rain was forecast for the Kaikoura coast, we always knew it was likely to turn up in Blenheim as a cloudy cool northeasterly with the odd shower.

But people dont think like that so much any more, the read a forecast and take everything in at fact. Example: A forecast says rain clearing in the morning and it clears at 3pm, that forecast was wrong. They dont look at it and say well the were close with the temp today Ill give them a 80% for todays forecast, they had an afternoon game of rugby and it rained. Getting the temp out +/- a couple of degrees is always going to happen, i think we all accept that. What I dont accept is the number of times metservice have stuffed a Severe Thunderstorm Outlook for Canterbury, it wasn't the models that was wrong in 3 cases I can think of it was the metservice forecasters. Maybe a better education of local conditions would improve the forecasts.

[Michael]

I remember when I was in Wanganui and it used to give our forecasts of the day as westerlies,either fine or with occasional showers,I am sure these days its the same as often seen a SW forecast there,SW actually are rare there and often turn up as a SE or westerly depending on its slant.NW wernt common though more often than a SW, either often as a drier W or as a northerly.