NOTICE!

The Weatherzone forum has now closed and is in read-only mode until the 1st of November when it will close permanently. We would like to thank everyone who has contributed over the past 18 years.

If you would like to continue the discussion you can follow us on Facebook and Twitter or participate in discussions at AusWeather or Ski.com.au forums.

Page 19 of 40 < 1 2 ... 17 18 19 20 21 ... 39 40 >
Topic Options
#1455470 - 02/03/2018 16:57 Re: QLD Tropical Cyclone Season 2017/2018 [Re: @_Yasified_shak]
Ken Kato Offline
Meteorological Motor Mouth

Registered: 05/03/2012
Posts: 6090
Originally Posted By: rainthisway
Originally Posted By: Ken Kato
Mangrovejack: Yeah I agree, waterlogged soil is always one of the things to consider when it comes to how easily certain types of trees topple over in strong winds.

1st image below is the forecast soil moisture change from the GFS over the next several days.

2nd image is the percentage of forecast members in the EC ensemble trying to spin up a TC between 7 and 9 days from now (pretty sure the "Tropical Storm" title uses the same categorisation of tropical system as the ECMWF themselves do
i.e. tropical storm = equivalent of Cat 1 TC in the Australian region):






Do you trust EC more then the other Models? Because i was reading somewhere, bloody wish I remembered the link, that overseas models are less reliable in Australia, due to there resolution over australia was lower then over there home ground as such. However your second pic has a resolution of 5km which is pretty awesome. So in your honest opinion which model do you believe is the most accurate more times than not.


Global models such as EC, GFS, etc have the same resolution around the world no matter what region.

There are various other models such as ACCESS and JMA just to name a couple which have different versions that focus on their home regional and local domains and these have finer horizontal resolutions than their own global version e.g. ACCESS-G covers the globe, ACCESS-R focuses on the Australian region and has a finer resolution, and ACCESS-C which covers capital city domains has an even finer resolution.

It's actually a popular myth that just because EC is a global overseas model that its resolution isn't as good as a home grown model such as ACCESS-R
e.g. EC's horizontal resolution of around 9km is actually better than that of ACCESS-R's ~12km (the 5km marked on the map is just referring to the map's resolution rather than the model itself).

It's also a popular myth that a global model is always less accurate than a home grown model. It depends on the model, how good its physics package is, and the way it assimilates observation data from many different sources. Resolution is very important but it's physics package and the way it assimilates obs data is also crucial.

As far as accuracy goes, a middle-of-the-ground multimodel consensus approach (with a slight bias towards the historically more accurate models) is more accurate (on average) than relying on any single model on its own.... even if it's EC. This is backed up by both objective skill stats as well as subjective assessments.

On average, EC has consistently been the most accurate model for decades for just about every weather variable at most levels of the atmosphere but that doesn't mean it beats the other models every time.
The best approach is to look at all of them, then use knowledge of TC principles and experience of what's happened in the past re forecasts vs observed to come up with the most likely scenario.

Top
#1455482 - 02/03/2018 18:27 Re: QLD Tropical Cyclone Season 2017/2018 [Re: Ken Kato]
MangroveJack70 Offline
Weather Freak

Registered: 01/02/2011
Posts: 229
Originally Posted By: Ken Kato
Originally Posted By: rainthisway
Originally Posted By: Ken Kato
Mangrovejack: Yeah I agree, waterlogged soil is always one of the things to consider when it comes to how easily certain types of trees topple over in strong winds.

1st image below is the forecast soil moisture change from the GFS over the next several days.

2nd image is the percentage of forecast members in the EC ensemble trying to spin up a TC between 7 and 9 days from now (pretty sure the "Tropical Storm" title uses the same categorisation of tropical system as the ECMWF themselves do
i.e. tropical storm = equivalent of Cat 1 TC in the Australian region):






Do you trust EC more then the other Models? Because i was reading somewhere, bloody wish I remembered the link, that overseas models are less reliable in Australia, due to there resolution over australia was lower then over there home ground as such. However your second pic has a resolution of 5km which is pretty awesome. So in your honest opinion which model do you believe is the most accurate more times than not.


Global models such as EC, GFS, etc have the same resolution around the world no matter what region.

There are various other models such as ACCESS and JMA just to name a couple which have different versions that focus on their home regional and local domains and these have finer horizontal resolutions than their own global version e.g. ACCESS-G covers the globe, ACCESS-R focuses on the Australian region and has a finer resolution, and ACCESS-C which covers capital city domains has an even finer resolution.

It's actually a popular myth that just because EC is a global overseas model that its resolution isn't as good as a home grown model such as ACCESS-R
e.g. EC's horizontal resolution of around 9km is actually better than that of ACCESS-R's ~12km (the 5km marked on the map is just referring to the map's resolution rather than the model itself).

It's also a popular myth that a global model is always less accurate than a home grown model. It depends on the model, how good its physics package is, and the way it assimilates observation data from many different sources. Resolution is very important but it's physics package and the way it assimilates obs data is also crucial.

As far as accuracy goes, a middle-of-the-ground multimodel consensus approach (with a slight bias towards the historically more accurate models) is more accurate (on average) than relying on any single model on its own.... even if it's EC. This is backed up by both objective skill stats as well as subjective assessments.

On average, EC has consistently been the most accurate model for decades for just about every weather variable at most levels of the atmosphere but that doesn't mean it beats the other models every time.
The best approach is to look at all of them, then use knowledge of TC principles and experience of what's happened in the past re forecasts vs observed to come up with the most likely scenario.



Thanks for the insights Ken. It sure is interesting watching the individual Models at the moment and I must agree, relying on one specific Model is not recommended. Grateful for the Charts you shared. I think we're about to enter quite an interesting little period with the MJO moving into our region in the next 2 or so weeks (approx). Wind Shear has been quite favourable of late and one would think with the Summer / Autumn change upon us, the cooler Antartic airflow may see the High pressure systems potentially pushing further up the Australian continent and with that, a more northerly push in SSTs and compressed +28 Deg C body of water through to the Equator.

Just taking a punt. Feel free to add your thoughts as I think we're all here to learn.


Thanks Ken.

Top
#1455490 - 02/03/2018 19:23 Re: QLD Tropical Cyclone Season 2017/2018 [Re: Ken Kato]
scott12 Offline
Weatherzone Addict

Registered: 10/11/2015
Posts: 1231
Loc: maadi Tully area
Originally Posted By: Ken Kato

The best approach is to look at all of them, then use knowledge of TC principles and experience of what's happened in the past re forecasts vs observed to come up with the most likely scenario.


Ken..That all sounds like what we know as an "Educated guess" really...

How often do we get a scenario where we can predict that a week out Cyclone Yasi is going to smash into Nth Queensland but there is so much uncertainty over most every other situation ..with so many people , resources and technology why is it so hard to predict whats going to happen a week away..why do we still have to make those guesses..?

Top
#1455500 - 02/03/2018 20:47 Re: QLD Tropical Cyclone Season 2017/2018 [Re: scott12]
Ken Kato Offline
Meteorological Motor Mouth

Registered: 05/03/2012
Posts: 6090
Originally Posted By: scott12
Originally Posted By: Ken Kato

The best approach is to look at all of them, then use knowledge of TC principles and experience of what's happened in the past re forecasts vs observed to come up with the most likely scenario.


Ken..That all sounds like what we know as an "Educated guess" really...

How often do we get a scenario where we can predict that a week out Cyclone Yasi is going to smash into Nth Queensland but there is so much uncertainty over most every other situation ..with so many people , resources and technology why is it so hard to predict whats going to happen a week away..why do we still have to make those guesses..?


Scott, "educated guess" has the connotation that every setup is virtually the same as using a dartboard when in reality, it's not.

It depends on the influences on a system's intensity and steering at the time. Many setups have high uncertainty while others low uncertainty. Yasi was a textbook example of the latter but there's also been many others where once the influences on a potential or existing TC become more clear-cut, its track and intensity have low uncertainty. You only need to look at many of the TC's that are about to get captured by an approaching upper trough to see how tight the consensus about its track becomes.
As another example of how a TC followed the general consensus of ensembles way out to 10 days, see the graphic below which I posted earlier in this thread - it's for Gita and the coloured tracks are all the forecast scenarios while the black track is what actually happened. The models were run way back on the 9th Feb only a short time after Gita was born. Considering the models had to forecast out to 10 days and the fact that it also had to forecast Gita's complete 180 followed by another curve towards NZ, I wouldn't call that an educated guess.
There's countless other similar cases as well but there's also a number of cases where the forecast scenarios were going everywhere.

Also, my previous post said nothing about how far ahead we're looking into the future. Of course, most setups are going to have high uncertainty beyond several days lead time. My post was talking in general, including the forecasting of TC's out to only a few days. Generally speaking, the closer you get to the forecast time period, the less of an "educated guess" it becomes.

But there's some setups where if both single and multi model ensembles are in good agreement that a system will form and intensity in a particular area at a forecast lead time of several days, it often does eventuate in reality.

P.S. as for this potential upcoming system currently being discussed in here, whenever various models are trying to spin up something in the Gulf or western Coral Sea more than a few days in advance but there's disagreement on the position, I like to interpret it as a "some kind of tropical disturbance signal that could either eventuate in the Gulf or Coral Sea". Given the proximity of the Gulf to the northwest Coral Sea, it only takes a slight change in the setup for a potential system to form in the Gulf instead of the western Coral Sea, or vice versa, or even both.
It's the nature of the beast in the tropics.




Edited by Ken Kato (02/03/2018 20:57)
Edit Reason: added stuff

Top
#1455501 - 02/03/2018 21:13 Re: QLD Tropical Cyclone Season 2017/2018 [Re: @_Yasified_shak]
tsunami Offline
Weatherzone Addict

Registered: 06/12/2010
Posts: 1373
Loc: Wynnum SE Brisbane
Thanks ken
Very informative
_________________________
Wynnum SE Brisbane

Top
#1455504 - 02/03/2018 22:55 Re: QLD Tropical Cyclone Season 2017/2018 [Re: @_Yasified_shak]
gawain Offline
Weather Freak

Registered: 04/07/2007
Posts: 465
Loc: Highgate Hill Brisvegas
Thank you Ken

Top
#1455506 - 03/03/2018 02:56 Re: QLD Tropical Cyclone Season 2017/2018 [Re: Ken Kato]
MangroveJack70 Offline
Weather Freak

Registered: 01/02/2011
Posts: 229
Originally Posted By: Ken Kato
Originally Posted By: scott12
Originally Posted By: Ken Kato

The best approach is to look at all of them, then use knowledge of TC principles and experience of what's happened in the past re forecasts vs observed to come up with the most likely scenario.


Ken..That all sounds like what we know as an "Educated guess" really...

How often do we get a scenario where we can predict that a week out Cyclone Yasi is going to smash into Nth Queensland but there is so much uncertainty over most every other situation ..with so many people , resources and technology why is it so hard to predict whats going to happen a week away..why do we still have to make those guesses..?


Scott, "educated guess" has the connotation that every setup is virtually the same as using a dartboard when in reality, it's not.

It depends on the influences on a system's intensity and steering at the time. Many setups have high uncertainty while others low uncertainty. Yasi was a textbook example of the latter but there's also been many others where once the influences on a potential or existing TC become more clear-cut, its track and intensity have low uncertainty. You only need to look at many of the TC's that are about to get captured by an approaching upper trough to see how tight the consensus about its track becomes.
As another example of how a TC followed the general consensus of ensembles way out to 10 days, see the graphic below which I posted earlier in this thread - it's for Gita and the coloured tracks are all the forecast scenarios while the black track is what actually happened. The models were run way back on the 9th Feb only a short time after Gita was born. Considering the models had to forecast out to 10 days and the fact that it also had to forecast Gita's complete 180 followed by another curve towards NZ, I wouldn't call that an educated guess.
There's countless other similar cases as well but there's also a number of cases where the forecast scenarios were going everywhere.

Also, my previous post said nothing about how far ahead we're looking into the future. Of course, most setups are going to have high uncertainty beyond several days lead time. My post was talking in general, including the forecasting of TC's out to only a few days. Generally speaking, the closer you get to the forecast time period, the less of an "educated guess" it becomes.

But there's some setups where if both single and multi model ensembles are in good agreement that a system will form and intensity in a particular area at a forecast lead time of several days, it often does eventuate in reality.

P.S. as for this potential upcoming system currently being discussed in here, whenever various models are trying to spin up something in the Gulf or western Coral Sea more than a few days in advance but there's disagreement on the position, I like to interpret it as a "some kind of tropical disturbance signal that could either eventuate in the Gulf or Coral Sea". Given the proximity of the Gulf to the northwest Coral Sea, it only takes a slight change in the setup for a potential system to form in the Gulf instead of the western Coral Sea, or vice versa, or even both.
It's the nature of the beast in the tropics.






Well said Ken. Probabilistic Modelling at its best.

Top
#1455507 - 03/03/2018 04:49 Re: QLD Tropical Cyclone Season 2017/2018 [Re: @_Yasified_shak]
rainthisway Offline
Weather Freak

Registered: 09/11/2015
Posts: 810
Loc: Watsonville, Atherton Tablelan...
Thanks Ken...I am glad you cleared that up for me. Much appreciated.
_________________________
Nikko

Somewhere over the rainbow

Top
#1455526 - 03/03/2018 13:33 Re: QLD Tropical Cyclone Season 2017/2018 [Re: Ken Kato]
scott12 Offline
Weatherzone Addict

Registered: 10/11/2015
Posts: 1231
Loc: maadi Tully area
Originally Posted By: Ken Kato
I like to interpret it as a "some kind of tropical disturbance signal that could either eventuate in the Gulf or Coral Sea". Given the proximity of the Gulf to the northwest Coral Sea, it only takes a slight change in the setup for a potential system to form in the Gulf instead of the western Coral Sea, or vice versa, or even both.
It's the nature of the beast in the tropics.



Thanks Ken for the explanation..

In regard to your above quote..is the difficulty in predicting where or if a system spins up along the northern flank of the country ,because ,if all the elements are present for cyclogenisis (is this what you called your "tropical disturbance signal"..?..)then its possible for a cyclone or cyclones to literally form anywhere along that low pressure trough line or monsoon trough where those conditions or "signal" exist...or is that too simplistic..?..

Top
#1455557 - 03/03/2018 21:54 Re: QLD Tropical Cyclone Season 2017/2018 [Re: @_Yasified_shak]
rainthisway Offline
Weather Freak

Registered: 09/11/2015
Posts: 810
Loc: Watsonville, Atherton Tablelan...
Access is really wanting this GoC or CS low.
_________________________
Nikko

Somewhere over the rainbow

Top
#1455564 - 03/03/2018 23:23 Re: QLD Tropical Cyclone Season 2017/2018 [Re: @_Yasified_shak]
Mick10 Offline
Weatherzone Moderator

Registered: 02/11/2001
Posts: 25465
Loc: Kirwan, Townsville - NQld.
gfs and ec push west into the NT
_________________________
Kirwan, Townsville Nth Qld -
June 2019 total - 27.8mm (21.2mm)
July 2019 total - 0.0mm (14.9mm)
2019 Yearly total to date - 2049.9mm (1129mm)

Top
#1455580 - 04/03/2018 09:37 Re: QLD Tropical Cyclone Season 2017/2018 [Re: @_Yasified_shak]
rainthisway Offline
Weather Freak

Registered: 09/11/2015
Posts: 810
Loc: Watsonville, Atherton Tablelan...
Access has got a coast hugger whereas Ec now has a stationary low over arnham land.
_________________________
Nikko

Somewhere over the rainbow

Top
#1455582 - 04/03/2018 09:41 Re: QLD Tropical Cyclone Season 2017/2018 [Re: @_Yasified_shak]
rainthisway Offline
Weather Freak

Registered: 09/11/2015
Posts: 810
Loc: Watsonville, Atherton Tablelan...
Also water vapour shows a low in cs but satelite clouds doesnt show it
_________________________
Nikko

Somewhere over the rainbow

Top
#1455589 - 04/03/2018 10:17 Re: QLD Tropical Cyclone Season 2017/2018 [Re: @_Yasified_shak]
Raindammit Offline
Weatherzone Moderator

Registered: 14/02/2002
Posts: 13367
Loc: Townsville & Bilyana NQ
Latest GFS has a TC hitting the Gladstone area on the 15th.
_________________________
Belgian Gardens, Townsville NQ
Bilyana FNQ

Top
#1455594 - 04/03/2018 11:07 Re: QLD Tropical Cyclone Season 2017/2018 [Re: @_Yasified_shak]
Rossby Offline
Weather Freak

Registered: 03/02/2018
Posts: 71
Just a little food for thoughts a west tracker with a building, in ridge.



Edited by Rossby (04/03/2018 11:12)

Top
#1455599 - 04/03/2018 11:57 Re: QLD Tropical Cyclone Season 2017/2018 [Re: @_Yasified_shak]
mysteriousbrad Offline
Weatherzone Addict

Registered: 23/10/2007
Posts: 1879
Loc: Rockhampton QLD
That would be rather nasty for Capricorn coast area!

Top
#1455608 - 04/03/2018 13:12 Re: QLD Tropical Cyclone Season 2017/2018 [Re: @_Yasified_shak]
rainthisway Offline
Weather Freak

Registered: 09/11/2015
Posts: 810
Loc: Watsonville, Atherton Tablelan...
We need this low to effect us so ross dam can hit 100. Then everyone will be happy.
_________________________
Nikko

Somewhere over the rainbow

Top
#1455624 - 04/03/2018 14:26 Re: QLD Tropical Cyclone Season 2017/2018 [Re: rainthisway]
Nature's Fury Offline
Weatherzone Addict

Registered: 29/11/2009
Posts: 2545
Loc: Brisbane Western Suburbs
Access and GFS both going for good CS systems around the same time. SSTs are heating nicely along the east coast as well.

Top
#1455625 - 04/03/2018 14:28 Re: QLD Tropical Cyclone Season 2017/2018 [Re: scott12]
Ken Kato Offline
Meteorological Motor Mouth

Registered: 05/03/2012
Posts: 6090
Originally Posted By: scott12
Thanks Ken for the explanation..

In regard to your above quote..is the difficulty in predicting where or if a system spins up along the northern flank of the country ,because ,if all the elements are present for cyclogenisis (is this what you called your "tropical disturbance signal"..?..)then its possible for a cyclone or cyclones to literally form anywhere along that low pressure trough line or monsoon trough where those conditions or "signal" exist...or is that too simplistic..?..


Yep you nailed it scott.
Whenever there's a belt of general cyclonic vorticity stretched out across a monsoon trough, I find that a low or multiple embedded lows can form anywhere along that trough and they could become a TC anywhere along that trough (but only as long as other factors such as favourably low shear, etc cooperate).
An example I often notice is when a monsoon trough lies across northern waters with the "eastern end" of it lying over the western Coral Sea. Some (but not all) of the models might try to spin up a TC in the Gulf but on occasion, a TC ends up forming on the eastern end of the monsoon trough instead (where it doesn't have to "compete" with the vorticity surrounding other potential systems) with the system in the Gulf remaining a low.

Top
#1455632 - 04/03/2018 15:38 Re: QLD Tropical Cyclone Season 2017/2018 [Re: Raindammit]
ColdFront Offline
Meteorological Motor Mouth

Registered: 29/06/2008
Posts: 19046
Loc: The Beach.
Originally Posted By: Raindammit
Latest GFS has a TC hitting the Gladstone area on the 15th.


CMC is showing evolution of the same area around the Solomons and running a low the same direction as GFS ,just a day or so later and weaker. Grain of salt stuff. Simply too far away.

_________________________


Top
Page 19 of 40 < 1 2 ... 17 18 19 20 21 ... 39 40 >


Who's Online
0 registered (), 36 Guests and 4 Spiders online.
Key: Admin, Global Mod, Mod
Today's Birthdays
Evil Kitty, Jake, ka0tik_pengwin, Nullamanna WX, shaunandwendy, stormygirl
Forum Stats
29947 Members
32 Forums
24194 Topics
1529247 Posts

Max Online: 2985 @ 26/01/2019 12:05
Satellite Image