Forecast issues in the ACT

[Gavin O'Brien]

Hi all, 

Far too often the forecast for us has in it "slight chance of a shower",yet with the exception of a very localised heavy shower and thunder which passed across Kambah,Woden and south Canberra before crossing to the south of the Airport around 2.15 to 3 pm this afternoon, there has been no significant weather since the big storm (18mm )  a week ago. Interestingly we had no rain here at Gilmore from this afternoon's event! I was driving my better half home from school on 

Weston when the storm developed. The rain was very intense from Mawson to Wanniassa but I drove out of it by the time I passed Gowrie!

Looking at the forecast for the next few days we have "sunny with the slight chance of a shower"! on several days.

Each way bet or computer generated nonsense?

Cheers

Gavin 

[Bussy R]

I'm with you Gav. Computer generated stuff never delivers on the majority of times from what I can see.

 

Rutherglen 100% chance of 7mm or something and we get nothing.

 

Bussy

 

--
You received this message because you are subscribed to the Google Groups "austpacwx" group.
To unsubscribe from this group and stop receiving emails from it, send an email to austpacwx+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/austpacwx/CADaZApoFFmQdtXcP_DL5i50Z9KmvmEk3iKAGSgMC9qy4tS3t_w%40mail.gmail.com.

[Andrew Miskelly]

Hi all,

Yes, presenting forecasts intuitively in the age of model-based everything is something which frustrates and fascinates me.

Some local examples of your frustrations here in the NSW Southern Highlands are:

• Overnight drizzle occurs regularly with easterly winds, but it's never forecast as such. "Possible shower" at best, but showers are a different beast.
• The "possible thunderstorm" thing. Occasionally fair, but when there's an upper ridge to the west, southerly or south westerly steering and a convergence edging inland, it'll never come off here. Storms will trigger along the convergence and move north, staying clear of the stable maritime air.

Human forecasters would (used to) pick these things.

On the other hand, I know the BoM's precip range forecasts drive people nuts, but in a sense they're quite objective.

An example was Brisbane last week. With Alfred approaching, the range for one day was 7 - 90 mm. That seems odd to 99% of people, but it accurately reflects the fact that the track and timing were greatly uncertain, and that the model spread was consequently large. Thus the 25th and 75th percentile values of 7 and 90 mm.

How do you find the line between sitting on the fence or having a bet each way, and transparently conveying uncertainty when it exists (which is most of the time)? That's the 64 thousand dollar question.

Andrew

To view this discussion visit https://groups.google.com/d/msgid/austpacwx/1UkI7QIPlI.1FPzEOgk9NX%40desktop-24tq0bb.