The myth of forecasting...?

[Dusk1983]

hi everyone, long time lurker, first time poster.

 

I'll keep this brief - things that've been bothering me for a while.

 

If a 5 day forecast changes every day - as they often do - then it's not a 5 day forecast. It's a 1 day forecast.

 

What use is a 1 day forecast to the general population, who can have a look outside and make a decent guess themselves? Not much. This Sunday's weather in my area has just been downgraded - 36 hours ago it was looking like 20C+ and plenty of Sun. Now we're heading for c. 15C and cloud. So what was billed as a 5 day forecast is in fact a 3 day forecast, at best. After all, it may well change again!

 

It seems that many forecasters live permanently in the future.

 

I've seen a page on model verification statistics but there appears to be suprisingly little interest in them. Some forecasters prefer to keep making predictions about the future without ever putting those to the test in the here and now  - not to mention being able to change them every day, if needed.

 

What are your thoughts?

[hillbilly]

hi everyone, long time lurker, first time poster.

 

I'll keep this brief - things that've been bothering me for a while.

 

If a 5 day forecast changes every day - as they often do - then it's not a 5 day forecast. It's a 1 day forecast.

 

What use is a 1 day forecast to the general population, who can have a look outside and make a decent guess themselves? Not much. This Sunday's weather in my area has just been downgraded - 36 hours ago it was looking like 20C+ and plenty of Sun. Now we're heading for c. 15C and cloud. So what was billed as a 5 day forecast is in fact a 3 day forecast, at best. After all, it may well change again!

 

It seems that many forecasters live permanently in the future.

 

I've seen a page on model verification statistics but there appears to be suprisingly little interest in them. Some forecasters prefer to keep making predictions about the future without ever putting those to the test in the here and now  - not to mention being able to change them every day, if needed.

 

What are your thoughts?

 

i used to be a farmer and a builder and relied on a good forcast and sometimes the best forcast is to open the curtains in a morning and look out.where i am in the hills i cannot count how many times we have been ready to bale a field for hay and given a nice sunny dayforcast and it has been misty or even drizzling all day,then again its harder to forcast in the hills just 10 miles away it can be as they say!

[Dusk1983]

I'm bumping this despite the general lack of interest (willingness?) to engage with the topic.

Another significant downgrade today (bank holiday monday) this time with the forecast changing OVERNIGHT from sunny to cloudy in the London area.

so, basically, no forecasting even happened. there was no forecast. No actual prediction, just changing the symbols according to the conditions out of the window. everyone may as well pack up and go home.

[Bobby]

Forecasting the weather is very difficult. The atmosphere is a constant state of flux and is almost infinitely complex, it's all about making a best guess with the data available at the time. As the data changes the forecast will adjust in order to get the most accuracy. A forecast for 5 days out is a best guess using the data available as to what the weather conditions will be like at that time. It's not a definitive statement as to what the weather will be like in 5 days time.  Forecasting accuracy has improved a huge amount mainly due to the latest supercomputer weather models, the 1 or 2 day forecast a few decades ago was as accurate as a 5 day forecast today.

 

No weather forecast is, ever has been or ever will be perfectly accurate due to the immense complexity and chaos of the atmosphere, not even 1 hour ahead. The weather forecast is always wrong, it's how wrong and it's much less wrong today than it ever has been.

 

Don't understand what you're getting at or what point your trying to make really.

Edited May 6, 2013 by Bobby