Met office - wrong

[Zod]

Sorry, but the title made me think of this:

https://www.youtube.com/watch?v=ReJgCZ0DGnY

[Moonhawk]

Gandahar said:

Unless you are having a wedding or judging on a life and death experience does it actually matter if right or wrong?

It matters to me at the moment because I have the roof off my house and it's only protected by a temporary scaffold, which is letting some water in.

An accurate forecast can mean the difference between me having to stay up all night on standby ready to start emptying buckets and moving tarps - compared to getting a decent night's sleep.

[mybrainhurts]

Moonhawk said:

Gandahar said:

Unless you are having a wedding or judging on a life and death experience does it actually matter if right or wrong?

It matters to me at the moment because I have the roof off my house and it's only protected by a temporary scaffold, which is letting some water in.

An accurate forecast can mean the difference between me having to stay up all night on standby ready to start emptying buckets and moving tarps - compared to getting a decent night's sleep.

And we're paying for the forecasts, so not getting value for money.

[Beati Dogu]

One solution:

https://www.youtube.com/watch?v=Dqw2KOOzdzQ

[LordGrover]

It's easier in sunny climates: youtube.

[ninja-lewis]

Moonhawk said:

Gandahar said:

Unless you are having a wedding or judging on a life and death experience does it actually matter if right or wrong?

It matters to me at the moment because I have the roof off my house and it's only protected by a temporary scaffold, which is letting some water in.

An accurate forecast can mean the difference between me having to stay up all night on standby ready to start emptying buckets and moving tarps - compared to getting a decent night's sleep.

The Met Office high resolution model uses a grid of 1.5km squares, which is still fairly large, both in terms of capturing the full detail of convection activity and forecasting which house is going to be hit by rain. It can easily be torrential rain in one street and dry a couple of streets over.

http://www.metoffice.gov.uk/research/news/2016/sci...

Compound that by most people obtaining a forecast through entering a postcode or town in a website/app or seeing a TV map symbol over an entire county. For example, probability of precipitation is given for around 7,000 locations in the UK, which is going to be relatively high level compared to most people's expectation of geographic resolution.

And of course most people don't understand how probability works. How many people would complain about a forecast 10% probability of precipitation being wrong just because it rained? If the same forecast conditions results in rain 1 in 10 times then that would actually be an accurate forecast.

http://www.metoffice.gov.uk/news/in-depth/science-...

So forecasting works if you're launching a large scale event (say D-Day) but you're on a hiding to nothing if you expect it to predict whether rain will spoil your BBQ. Looking at a rain radar will give a better short term idea of what is happening if you need that level of precision.

[Tempest_5]

At the turn of the century I worked with a Meteorological organisation for a few years. Speaking to the the meteorologists I got the impression anything more than 4 to 5 days in advance was getting into the realms of speculation. One thing they were, and still are, doing is gathering data and refining their prediction models. This has added to the improved forecasting.

If you are interested in weather there are some good books out there written for pilots which give a good overview without going into PHD level theory.