The joke that is weather forecasting (3 Dec, 2018)

Have your say on today's Aardvark Daily column

The joke that is weather forecasting (3 Dec, 2018)

Postby aardvark_admin » Mon Dec 03, 2018 8:58 am

This column is archived at: https://aardvark.co.nz/daily/2018/1203.shtml

How could NIWA and MetService get their seasonal outlooks so wrong for the past month or so?

Why are my legs still lilly-white and covered in denim instead of slightly browned and protruding from my big-boy shorts, as they usually are by this time of the year?

Why haven't NIWA's new supercomputers seemingly done anything to improve the accuracey of their forecasts and outlooks?

Is there any point in publishing a prediction that might have just a 45% confidence factor associated with it?
aardvark_admin
Site Admin
 
Posts: 3538
Joined: Wed May 07, 2014 2:10 pm

Re: The joke that is weather forecasting (3 Dec, 2018)

Postby hagfish » Mon Dec 03, 2018 9:21 am

My go-to for weather forecasts - particularly if I'll be in the hills - is yr.no. If found their accuracy to be pretty good.

As for reliable forecasts; if there is a huge (anti) cyclone on its way, then we can be reasonably sure about what the next day or two will bring.

For every other day, why can't they just say, "meh - be sure to apply sunscreen, but remember your raincoat".
hagfish
 
Posts: 717
Joined: Thu May 08, 2014 10:28 am

Re: The joke that is weather forecasting (3 Dec, 2018)

Postby GSVNoFixedAbode » Mon Dec 03, 2018 9:31 am

I'd subscribed to a Snow Alert ('snowmail') for a few of the Otago skifields some time ago. It's a European-based system that gives a 3-day warning of heavy weather, with snow & freezing altitudes. What I've seen over the past 3+ years is that the accuracy of their system was pretty much on the button, while local forecasting was a bit hit&miss.

For local weather I've got a vague memory of years past when coastal & other watchers called in their observations, which were then used to put together a forecast. I suspect that was more accurate for short-term forecasts but the super-silicon soothsayer gives a better long-term (10-14 days) forecast these days. Predicting the weather for the next few months? That still seems to be just educated guesswork.
GSVNoFixedAbode
 
Posts: 251
Joined: Thu May 08, 2014 8:53 am

Re: The joke that is weather forecasting (3 Dec, 2018)

Postby aardvark_admin » Mon Dec 03, 2018 9:34 am

Or...
Light to strong nor-southerlies with the possibility of an east-westerly shift. Potential sunny periods with possible light rain and a chance of heavy downpours. Temperatures variable or constant around normal except when colder or warmer than normal

I think that's accurate enough eh?

And yes... I think shifting to a reliance on automation and computer models rather than some "experienced interpolation" by observers around the country has not improved the accuracy of forecasting. I do notice too many days when the forecast is completely at odds with the weather I see outside my window.
aardvark_admin
Site Admin
 
Posts: 3538
Joined: Wed May 07, 2014 2:10 pm

Re: The joke that is weather forecasting (3 Dec, 2018)

Postby goosemoose » Mon Dec 03, 2018 9:34 am

Where i live I generally expect the opposite of what NIWA predicts. I also like the daily Metservice rain forecast for our area that predict 40-60 mils of rain in 4-6 hour period a day or 2 in advance then peter out to 0.2 mils as you get closer to the predicted "downpour".

Though in all fairness to them they are "generally" accurate as in helping me choose what clothes to wear that day!
goosemoose
 
Posts: 581
Joined: Thu May 08, 2014 1:05 pm

Re: The joke that is weather forecasting (3 Dec, 2018)

Postby latewings » Mon Dec 03, 2018 9:59 am

Metservice is no longer a forecasting service, it's a reporting service. All it tells us is what the weather is currently doing. For anything aviation related you have more chance smelling a fart in a hurricane than getting accurate metservice data.

Most flytypes use metvuw.com, the accuracy of their information is far greater. You have to interpret a little, but it's not too difficult.
latewings
 
Posts: 349
Joined: Thu May 08, 2014 10:18 am

Re: The joke that is weather forecasting (3 Dec, 2018)

Postby hagfish » Mon Dec 03, 2018 10:06 am

latewings wrote:Metservice is no longer a forecasting service, it's a reporting service. All it tells us is what the weather is currently doing. For anything aviation related you have more chance smelling a fart in a hurricane than getting accurate metservice data.

Most flytypes use metvuw.com, the accuracy of their information is far greater. You have to interpret a little, but it's not too difficult.


I use metservice's rain radar all the time. I can usually extrapolate from that if it's safe to go for a walk. But that's about it. One thing I find quite 'telling' is that they report the current surface pressure, but not the previous few hours' pressures. Knowing the one reading in isolation is next to useless, imo.
hagfish
 
Posts: 717
Joined: Thu May 08, 2014 10:28 am

Re: The joke that is weather forecasting (3 Dec, 2018)

Postby Stevesub1 » Mon Dec 03, 2018 10:48 am

Weather has been crap, last week in Wellington, cold and wet, now home in the far north warm and wet!!!! Forecasting the weather in NZ is crap. I now revert to the Windy app that seems to be the best. I rely on the Marine forecasts especially for wind, well I do better looking out the window at the local hardware shops flag. I have missed fishing days due to the forecast saying there will be lots of wind when there is none and other times I have gone out when there is supposed to be no forecast wind for the day and up it comes a hour or so after we are out there. It can be quite dangerous if the forecast is wrong for fisherman in small boats like us. Come on Metservice, you have a reasonable web site for marine forecasts, just get it right or nearer to right than it actually is right now. Our safety depends on it.
Stevesub1
 
Posts: 25
Joined: Wed Sep 26, 2018 10:24 am

Re: The joke that is weather forecasting (3 Dec, 2018)

Postby granada29 » Mon Dec 03, 2018 10:49 am

My own method of forecasting is to predict that the weather tomorrow will be the same as it is today. That seems to be a more accurate method than believing the forecasts that I read in the newspapers. Rainy days (sunny days) are like bannanas, they come in bunches.
granada29
 
Posts: 31
Joined: Thu Sep 29, 2016 8:38 am

Re: The joke that is weather forecasting (3 Dec, 2018)

Postby Kiwiiano » Mon Dec 03, 2018 11:16 am

I checked out MetVUW a few years back by clipping their 10 day prediction then following it day by day until I saw what actually happened. It was for Nth Tasman sea depression to track SE and pass across Cook Strait. The refined predictions vacillated a bit during the 10 days, uncertain about whether it may move further eastward but on day zero it went smack thru the Straits .
The current gloom, despond and falls of toads may well have an element of climate change if the Southern Hemisphere is getting the big blocking highs that are a curse in the Northern Hemisphere.
~ Kiwiiano
“I'm right 98% of the time, so who gives a damn about the other 3%?"
Kiwiiano
 
Posts: 441
Joined: Wed May 07, 2014 5:36 pm

Next

Return to Today's column

Who is online

Users browsing this forum: No registered users and 4 guests