Page 1362 of 1363 < 1 2 ... 1360 1361 1362 1363 >
Topic Options
#1407251 - 12/02/2017 12:58 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
ThD Ht Offline
Weatherzone ratbag

Registered: 02/03/2002
Posts: 3775
Loc: Reynella
Originally Posted By: petethemoskeet
Hey Galaxyman tell us how you really feel laugh grin

he can't do that pete, the profanity will get him banned.
don't worry Gm, there must be a hot rock over there somewhere
you can crawl under and bake until the next skin blistering
load of tripe comes along.
the wind is cool but when the sun comes out from behind the
clouds it still bites.
got the doors wide open and a stiff breeze outside, nice to
get a change through the place for some fresh air so i can
get some fresh air through the place for a change.

TH
_________________________
Welcome to South Australia,
the third world basket case state.

Top
#1407257 - 12/02/2017 13:07 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
-Cosmic- (naz) Offline
Meteorological Motor Mouth

Registered: 27/08/2003
Posts: 6674
Loc: Adelaide Hills, SA
Beautiful refreshing change signals end of the hot spell smile .

Top
#1407271 - 12/02/2017 13:20 Re: South Australian and Adelaide day to day forecasting [Re: petethemoskeet]
Galaxyman Offline
Weather Freak

Registered: 30/01/2016
Posts: 174
Loc: Renmark
Originally Posted By: petethemoskeet
Hey Galaxyman tell us how you really feel laugh grin I'll swap our weather with ya,love some cool stuff for a change.


Sadly I can't because this is a PG rated forum. I have a few choice 4-letter words and gestures for this rancid tripe. sick mad

I wouldn't care if it was a more sensible cool down that lasts for a couple days, but there is no excuse for this polar attack and the absurd disgrace tsunami of antarctic that awaits us on Feb 17.


Top
#1407380 - 12/02/2017 17:13 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
rstewart84 Offline
Weatherzone Addict

Registered: 17/01/2007
Posts: 2036
Loc: Kalangadoo,South Australia
Calm down Galaxyman whilst this hot spell maybe over if what a mate of mine has told me we will only have to wait 10 months till we get thrown in the oven by the extreme heat whilst NSW suffers from the polar attack and tsunami of antarctic all summer long

I am pretty sure Thunda can back me up on this as were all very observant of the weather around Australia but is welcome to correct me being the wise Jedi master he is wink from what i have seen over the last few months is a distinct lack in TC's up in QLD and WA that have formed and made landfall

I'm pretty sure this can only mean a sign of an El Nino forming shortly when i don't know but apparently the current Indices for either IOD or Enso not sure which one my mate mentioned to me over the phone is currently sitting at -2 for it to be a La Nina it has to be between 0-7

But thats just my take on it as i mentioned above Thunda is welcomed to clarify and correct it


Top
#1407405 - 12/02/2017 19:21 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
ThD Ht Offline
Weatherzone ratbag

Registered: 02/03/2002
Posts: 3775
Loc: Reynella
hi RS, jeeze you've dropped me in it now mate....thx.
your right about TCs, quiet as but the IOD has been doing
the trick regardless of TCs or not, huge gobs of moisture
coming down through W.A., you could just about start
cropping and bailing in central W.A. at the moment.
an amazing year over there. now the sw corner is getting
drilled. you have to remember we get our weather from three
different directions, from the sw with cold fronts, from
the w/nw as a result of moisture feeding down from nw W.A
and to a lesser extent moisture fed down from the eastern
states. el nino doesnt have the effect on us like it does
the ne quadrant of australia and the eastern states in
general, because if it craps out we have 2 other backups.
the interaction between the three and when they peak and
trough in relation to each other and the season and time
of year is complex but understandable. way too much for
me to go into here. SOI has been running +ve but is
yo-yoing around 0 at the moment sitting on the fence so
could trend either way. currently -2 but has been running
in the +ve.
TS is far more qualified to clarify on this, he's a pro
i'm just an amatuer.

TH
_________________________
Welcome to South Australia,
the third world basket case state.

Top
#1407420 - 12/02/2017 21:29 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
teckert Offline
Weatherzone Moderator

Registered: 27/05/2001
Posts: 17361
Loc: NE suburbs, Adelaide, South Au...
no sign of a return of el nino. However the liklihood of a stronger La Nina is on the cards later this year or into next.

Top
#1407422 - 12/02/2017 21:36 Re: South Australian and Adelaide day to day forecasting [Re: teckert]
Eevo Offline
Weatherzone Addict

Registered: 02/11/2013
Posts: 1492
Loc: Bridgewater
Originally Posted By: teckert
no sign of a return of el nino. However the liklihood of a stronger La Nina is on the cards later this year or into next.


so wet weather for eastern and south australia.

Top
#1407423 - 12/02/2017 21:45 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
teckert Offline
Weatherzone Moderator

Registered: 27/05/2001
Posts: 17361
Loc: NE suburbs, Adelaide, South Au...
more than likely, but appears to be a slightly drier autumn and winter, and possibly spring to come first.

Top
#1407872 - 15/02/2017 22:11 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
Chris #3 Offline
Weatherzone Addict

Registered: 08/02/2009
Posts: 1862
Loc: Semaphore SA
Originally Posted By: -Cosmic- (naz)

It probably doesn't matter that much anymore because the knowledge base for understanding how the weather works is far greater than any 5- or 10-minute news spiel could capture in one news segment. I've invested enough time and effort to get my own [private] ideas/models of how the weather works up and running smile .


Naz got me thinking.

Do you guys ever go back and test your forecasts?
Was pondering this the other day and surprised more statistical analysis isn't done on accuracy (for WZ/BOM/different models)

You could make some fancy graphs and see how our forecasting has improves through time.


Originally Posted By: Chris #3
Originally Posted By: Jet entrance
Given what I'm seeing on latest model runs, tomorrows temp will be bumped up across the board by 2 or 3 deg.

I'd say 43 or 44 for Adelaide


I don't think it'll quite make the forecast. Just shy of 42 for me - although it'll feel just like it or worse.
(used meteoearth, the dragonflies and my gutfeel for my prediction- the Chris model is always meteoearth - gutfeel and an animal)


Chris 2017 forecast review -
8th Feb Adelaide
Forecast: Just shy of 42c
Actual maximum temperature: 41c (no decimal points with the records I'm using frown )

Hrm how would I measure my accuracy. Percentage?
Predicted probability of temperature/rain
Against
Observed temp/rain/whatever.

Or comparison against BOM?

Anyways -
Chris #3 outforecasted the WZ SA group.

***Best in Show - forecasting 8th Feb - Chris #3***
Forecast derived from: MeteoEarth, Dragonflies, Gut feel.

Yeah - a few more of these and I'll be thinking I'm far better at forecasting the weather than the BOM. Confirmation bias is so flattering smile

PS - not a dig at you naz. You are beyond my reach knowledge-wise.



Edited by Chris #3 (15/02/2017 22:15)
_________________________
GFS Stormcast by BSCH




Top
#1407876 - 15/02/2017 22:59 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
Markus Offline
Weatherzone Moderator

Registered: 02/12/2010
Posts: 2015
Loc: Clare, SA
Not directly comparable, but I find looking back over my own forecasts (in my case, it's generally only 'storm events' I really care about, rather than day to day thus having many variables to account for), but I would suggest I learn 80% of what I know now from the post analysis of events, it is astonishing how many little things can go unnoticed in models and in the outcome without looking back over it all. You start noticing patterns in models and outcomes, local topographic features affecting weather etc. All of these observations tend to improve ones forecasting capabilities over time.

Using the September tornado outbreak as an example, near Port Broughton I was aware that there were multiple mesocyclones along a flanking line, but it wasn't until I looked back through my images, hours of watching radar loops etc did I realise I had watched an updraught replacement cycle occur. Without the post analysis I likely would have never realised.

Another example is a day I chased out onto the flats near Morgan, only for very severe storms to form on the ranges far to my west...lesson learned that day, ensure there's an adequate trigger in the target area lol. (haven't made a similar mistake since....yet!)

I think for the big models, e.g. EC, they frequently test old versions against new ones to see how forecast accuracy has improved in certain areas. Often you'll see improvements in one area, but significant regression in others.


More specific to your question Chris, I reckon a running graph where you plot how many degrees up or down from the forecasted temp vs observed, along a line would visualise temp forecasting nicely.
The straighter the line and the closer to 0, the better the forecasting!
_________________________
My Blogging site.

http://markdawsonphoto.wordpress.com/

Top
#1408005 - 16/02/2017 20:46 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
-Cosmic- (naz) Offline
Meteorological Motor Mouth

Registered: 27/08/2003
Posts: 6674
Loc: Adelaide Hills, SA
Originally Posted By: Chris #3
Originally Posted By: -Cosmic- (naz)

It probably doesn't matter that much anymore because the knowledge base for understanding how the weather works is far greater than any 5- or 10-minute news spiel could capture in one news segment. I've invested enough time and effort to get my own [private] ideas/models of how the weather works up and running smile .


Naz got me thinking.

Do you guys ever go back and test your forecasts?
Was pondering this the other day and surprised more statistical analysis isn't done on accuracy (for WZ/BOM/different models)

You could make some fancy graphs and see how our forecasting has improves through time.

Hi smile ,

Unfortunately, I dont have access to a lot of government-based data that could help with better understanding improvements in forecasting skill. However, there is a graph already put together in the Reducing Vulnerability to Weather and Climate Extremes book, available from the Bureau. On page 14 it shows how computer power and model skill has improved, 1971 to 2001, for a 24-hour forecast, Australian Region. Computer power improved from less than 0.01 Gigaflops to (in 2001) over 100. The model skill index changed from useful to half-way between useful and excellent, although this does not factor in 2002-2016, its still 30 years. The biggest change in skill occurred sometime 1995-1998.

My model is based on a control period from about mid-May 2012 to the end of 2016, on the daily and sub-daily time-scale. It is very local or area specific (based on local and regional dynamics more than global), so unfortunately, apart from local weather stations I dont have much to compare it to (and yes I have done comparisons shorter term variability is not easy, medium-term is reasonable).

Originally Posted By: Chris #3
Hrm how would I measure my accuracy. Percentage?
Predicted probability of temperature/rain
Against
Observed temp/rain/whatever.

Or comparison against BOM?

Root Mean Square Error (RMSE), and maybe the P-value (from regression modelling).

Top
#1408040 - 17/02/2017 06:28 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
ThD Ht Offline
Weatherzone ratbag

Registered: 02/03/2002
Posts: 3775
Loc: Reynella
the improvements are not just down to computing power.
all the computing power in the world is useless without
the algorithms that drive them, for the people who write
the code to produce these model outputs they are constantly
reviewing and comparing real time analysis and past
prognostic model outputs to work out where the programs can
be tweaked to bring them closer to match final real world
analysis.
there is also the matter of data input, without reliable,
accurate and timely data delivered to said computer all the
above is irrelevant, over the time frame they're talking about
there has been a massive improvement in both the quantity and
quality of data available although in relation to computer
power vs. data available it's always been computer power
that's been playing catchup. computer power whilst essential,
itself has been more of a secondary aide to model improvement
allowing the running of complex programs and crunching of huge
amounts of data. even a (comparatively speaking) modest
mainframe computer can run the algorithms and data to arrive
at an output, it will just take years to do it. the computing
power allows for timely deliverance of an output.
the book seems to indicate to the layman that "computing power
alone is responsible for "model skill"improvement although for
those more aware of how things work "model skill" could be
interpreted as the software.
in the end all three are essential to each other.
100 gigaflops is a tad old too, past the tera flop stage
with the top 10 supercomputers running well into peta flops,
a peta flop being 1 million giga flops.
----------
QUOTE:-
A new supercomputer from China has topped the latest list of
the world's most powerful machines. The 93 petaflop Sunway
TaihuLight is installed at the National Supercomputing Centre
in Wuxi. At its peak, the computer can perform around 93,000
trillion calculations per second.
----------
that's almost 1 million times faster than the 100 gigaflop
machine back in 2001.
i actually noticed a large improvement in model accuracy
around 2007-2010 i think, not sure if it was real or imagined.

TH
_________________________
Welcome to South Australia,
the third world basket case state.

Top
#1408222 - 17/02/2017 20:34 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
Jet entrance Offline
Weather Freak

Registered: 16/12/2015
Posts: 409
Loc: Hallet Cove, SA
Big shower here just now at Noarlunga, very heavy but the only one btl's.

Top
#1408230 - 17/02/2017 21:16 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
-Cosmic- (naz) Offline
Meteorological Motor Mouth

Registered: 27/08/2003
Posts: 6674
Loc: Adelaide Hills, SA
Ok smile , here comes some potential interesting shower activity (in A/Hills).

Top
#1408493 - 19/02/2017 06:28 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
ThD Ht Offline
Weatherzone ratbag

Registered: 02/03/2002
Posts: 3775
Loc: Reynella
soooooooo........the weather is about as exciting as a 3hr
lecture on the reproductive habits of african nematodes.
so what happens when the views reach 100 million, is there
space in the window for the extra digit or does it reset to 0?

oh, nearly forgot, did actually get lacquered by one decent
heavy shower yesterday around 3:45 on the southern expressway
going home after catching up with TS. unfortunately it didn't
translate into anything significant where i live with just a
few spots barely wetting the concrete and the only thing that
dribbled into the rain guage was me.
things are wetish outside, radar has drizzle showers going
through adelaide and surrounds most of the night.
today is looking like it will turn out to be a day of serious
full on hardcore heavy duty anticyclonic gloom and more
scattered drizzle.




TH
_________________________
Welcome to South Australia,
the third world basket case state.

Top
#1408524 - 19/02/2017 10:40 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
ThD Ht Offline
Weatherzone ratbag

Registered: 02/03/2002
Posts: 3775
Loc: Reynella
wow, got a real dose of winter out there,
almost constant drizzle-medium shower for
the last hour or so accompanied with a light
to stiff breeze.
not exactly the anticyclonic gloom it appeared
on layers, more cellular and showery in nature.
wouldnt be surprised if these stream showers
intensify and a bit more convective this arvo
although the ranges and southeast will prolly
get the better of it.
4.2mm all up so far.

TH
_________________________
Welcome to South Australia,
the third world basket case state.

Top
#1408527 - 19/02/2017 10:46 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
Galaxyman Offline
Weather Freak

Registered: 30/01/2016
Posts: 174
Loc: Renmark
Sounds depressing as hell. At least here there is no drizzle and the sun has been out with partly cloudy stratocrapulus rather than the slate-grey nuclear winter that places like southern Victoria are famed for.

Top
#1408534 - 19/02/2017 11:30 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
Pooley72 Offline
Cloud Gazer

Registered: 17/01/2017
Posts: 8
Loc: Seacliff Park
7.2mm at Seacliff Park since midnight. Adds to the 25.6 we had already had for a MTD of 32.8. Unbelievable summer weather...
_________________________
My PWS: www.wunderground.com/personal-weather-station/dashboard?ID=INOARLUN2

Jan: 52.2mm
Feb: 32.8mm
Mar:
Apr:
May:
Jun:
Jul:
Aug
Sep:
Oct:
Nov:
Dec:

2017: 85.0mm

Top
#1408537 - 19/02/2017 11:41 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
StormCapture Offline
Weatherzone Addict

Registered: 19/12/2013
Posts: 1557
Loc: Uleybury
Mere 2.4mm over the two days
42.4 MTD, 115.6YTD

Top
#1408598 - 19/02/2017 15:48 Re: South Australian and Adelaide day to day forecasting [Re: Thunderstruck]
red earth Offline
Weather Freak

Registered: 21/12/2007
Posts: 355
Loc: Lockleys, Adelaide
Pretty challenging in the gusty wind at West Lakes today for our dragon boat racing. No rain however.

Top
Page 1362 of 1363 < 1 2 ... 1360 1361 1362 1363 >


Moderator:  Helen, Lindsay Knowles, Markus, teckert 
Who's Online
1 registered (Popeye), 273 Guests and 5 Spiders online.
Key: Admin, Global Mod, Mod
Today's Birthdays
Sako338
Forum Stats
29077 Members
32 Forums
23498 Topics
1436949 Posts

Max Online: 2925 @ 02/02/2011 22:23
Satellite Image