OK...Time for a Storm2K Concensus Forecast...Please Join

#1 Postby **MWatkins** » Fri Sep 17, 2004 9:32 pm

The sum of us is smarter than any one of us right? At least that's how concensus forecasting works. So...with this in mind why don't we get a Storm2K Ensemble going?

Here's how it work. Twice per day...from now until Jeanne goes away (and we could revive this for any storm)...I would like to complie forecasts for Jeanne here in this post and see how our tracks look together.

Here's what's needed from you, the forecaster...or ANYBODY who wants to take a crack at this.

No later than 10:30AM or 10:30PM respectively, please post, in this thread, your track forecast for Jeanne (note this is 1/2 hour before the NHC product comes out).

You will need to forecast position and intensity for the following times for the 10:30AM track, starting with the NHC position at 8AM:

Initial - 8AM
12 Hours - Today at 8PM
24 Hours - Tonight at 8PM
36 Hours - Tomorrow at 8AM
48 Hours - Tomorrow at 8PM
72 Hours - Day 3 at 8AM
96 Hours - Day 4 at 8AM
120 Hours - Day 5 at 8AM

You get the point (for the ...some folks are doing this already....so I can lift your forecast but it would be cooler if you posted it here. I'll do the same. Include the discussion if you want...but really need the track and the projected intensity.

From there...I will complie the tracks and provide a concensus track for all members who submitted a track...and then verify numbers to see how our forecasts fare.

I'll give an example using my forecast for tonight...which is just me lifting the concensus forecast for tonight from the main model guidance:

Initial: 8PM 20.2N 72.0W 40 Knots
12 hours: 21.5N 73.0W 35 Knots
+24: 22.9N 73.4W 35 Knots
+36: 24.4N 73.4W 40 Knots
+48: 26.2N 73.0W 45 Knots
+72: 29.2N 71.5W 50 Knots
+96: 28.6N 71.3W 50 Knots
+120: 27.0N 73.3W 55 Knots

Come on...give it a try. FB...Stormsfury...NWxWiz and others...this means you. And ANYBODY else...doesn't matter how bad you think your forecast is...Let's see what we can do.

MW

huricanwatcher · #2 Postby **huricanwatcher** » Fri Sep 17, 2004 9:37 pm

interesting...... lets place our bets....

#3 Postby **MWatkins** » Fri Sep 17, 2004 9:41 pm

Forecast for 9/18/2004 00Z

Initial: 8PM 20.2N 72.0W 40 Knots
+12: 21.5N 73.0W 35 Knots
+24: 23.0N 73.5W 35 Knots
+36: 24.5N 73.5W 40 Knots
+48: 26.0N 73.0W 45 Knots
+72: 29.0N 71.5W 50 Knots
+96: 28.5N 71.5W 50 Knots
+120: 27.0N 73.0W 55 Knots

huricanwatcher · #4 Postby **huricanwatcher** » Fri Sep 17, 2004 9:54 pm

right now 45 mph...... but looking like ....... if ya wanna catch this one, gonna need a big Hook........

COMEAR FISHY FISHY FISHY

#5 Postby **MysticOne** » Fri Sep 17, 2004 9:59 pm

Cool Idea! :idea:

SouthernWx · #6 Postby **SouthernWx** » Fri Sep 17, 2004 10:02 pm

Well, it's after 10:30 pm, but I created this forecast at 9 pm....after viewing the 00z data..

Forecast for 9/18/2004 00Z
Initial: 8PM 20.2N 72.0W 40 Knots

+12: 20.9N 73.1W 40 kts
+24: 22.0N 74.0W 35 kts
+36: 23.2N 74.2W 40 kts
+48: 24.5N 74.0W 45 kts
+72: 27.5N 72.0W 55 kts
+96: 28.0N 68.0W 60 kts
+120: 26.5N 66.5W 65 kts

It's loop-de-loop time for miss Jeanne

#7 Postby **MWatkins** » Fri Sep 17, 2004 10:11 pm

Members: 4 (added StormsFury)

MW
NWHCC
SouthernWx
StormsFury

Track and Intensity Results:

0: 20.2N 72.0W 40 knts
+12: 21.0N 73.0W 39 knts
+24: 22.1N 73.7W 39 knts
+36: 23.3N 74.1W 41 knts
+48: 24.9N 74.2W 44 knts
+72: 27.8N 73.7W 50 knts
+96: 28.5N 73.1W 54 knts
+120: 27.9N 73.7W 60 knts

MW

Stormsfury · #8 Postby **Stormsfury** » Fri Sep 17, 2004 10:17 pm

Oh man, it's 11:00 pm on the nose and I missed this thread altogether until now ...

Alright, here goes ...

Code: Select all

8 pm 9/17/04 -
000 - 20.2ºN, 72.0ºW, 40 kts ...
012 - 21.0ºN, 72.9ºW, 40 kts ...
024 - 22.0ºN, 73.6ºW, 45 kts ...
036 - 23.9ºN, 74.0ºW, 45 kts ...
048 - 25.9ºN, 74.1ºW, 45 kts ...
072 - 28.5ºN, 74.3ºW, 45 kts ...
096 - 29.4ºN, 75.0ºW, 50 kts ...
120 - 29.7ºN, 75.6ºW, 60 kts ...

#9 Postby **MWatkins** » Fri Sep 17, 2004 10:28 pm

Stormsfury wrote:Oh man, it's 11:00 pm on the nose and I missed this thread altogether until now ...

Alright, here goes ...

Code: Select all
8 pm 9/17/04 - 000 - 20.2ºN, 72.0ºW, 40 kts ... 012 - 21.0ºN, 72.9ºW, 40 kts ... 024 - 22.0ºN, 73.6ºW, 45 kts ... 036 - 23.9ºN, 74.0ºW, 45 kts ... 048 - 25.9ºN, 74.1ºW, 45 kts ... 072 - 28.5ºN, 74.3ºW, 45 kts ... 096 - 29.4ºN, 75.0ºW, 50 kts ... 120 - 29.7ºN, 75.6ºW, 60 kts ...

Got it in...please see updated post above and thanks for participating!
All tracks now closed until tomorrow's 10:30AM track.

MW

#10 Postby **JPmia** » Fri Sep 17, 2004 10:35 pm

You must have read that book on the accuracy of crowd decision making...good idea b/c it actually works!

ColdFront77 · #11 Postby **ColdFront77** » Fri Sep 17, 2004 10:47 pm

* sigh *

I am up to doing this as much as I am not.

I don't see how we wouldn't follow each others coordinates in some fashion.

#12 Postby **MWatkins** » Fri Sep 17, 2004 10:50 pm

ColdFront77 wrote:* sigh *

I am up to doing this as much as I am not.

I don't see how we wouldn't follow each others coordinates in some fashion.

I actually already have a database set up to handle the intake of the data and the aggregate data...so really this will be just for folks to put their tracks up. Perhaps rather than posting the track here I should start a new thread when the forecasts are finalized.

MW

MBryant · #13 Postby **MBryant** » Fri Sep 17, 2004 10:55 pm

Crowd decision making has a place, but this isn't one of them. I'll let the crowd determine the favorite, most likely, least liked and other subjective items. You need experts and knowledgable amatures to analyze systems.

I wouldn't want a crowd to tell me how to fix my car, but I want a mechanic, whether professional or amature. But I might want a crowd's opinion on what color they like for resale value.

Following the crowd only helps you survive the disasters in a corporate environment by allowing you to shift the responsibility to the crowd.

ColdFront77 · #14 Postby **ColdFront77** » Fri Sep 17, 2004 10:59 pm

There has to be a general idea to know which latitude, longitude and wind speed to even be moderately right.

What I am trying to say isn't coming out right, what else is new, huh?

#15 Postby **MWatkins** » Fri Sep 17, 2004 11:00 pm

MBryant wrote:Crowd decision making has a place, but this isn't one of them. I'll let the crowd determine the favorite, most likely, least liked and other subjective items. You need experts and knowledgable amatures to analyze systems.

I wouldn't want a crowd to tell me how to fix my car, but I want a mechanic, whether professional or amature. But I might want a crowd's opinion on what color they like for resale value.

Agreed to a point. So far the 4 tracks submitted this evening come from some of the most respected members of this board. So far (myself excluded) you have a very realistic and representative track from some very knowledgable folks.

If I really get busy I can figure out a way to weight tracks that are verifying better than others and create a true ensemble...but we'll have to see about that...that may be a 2005 goal.

MW

#16 Postby **MWatkins** » Fri Sep 17, 2004 11:02 pm

ColdFront77 wrote:There has to be a general idea to know which latitude, longitude and wind speed to even be moderately right.

What I am trying to say isn't coming out right, what else is new, huh?

I get you. Your point is...to even attempt a forecast like this you have to be knowledgable to start with...especially in terms of speed...course...direction and intensity. It is not....in other words...a popularity contest like American Idol.

MW

MBryant · #17 Postby **MBryant** » Fri Sep 17, 2004 11:11 pm

On second thought, you do have a hypothesis which you are attempting to test, so why not.

It's not like your asking anyone to use this result in place of the ones from the experts on this board and at the NHC. I may have overreacted.

Besides, as others have noted, only somewhat knowledgable people would take the time to respond in this format. So go for it.

ColdFront77 · #18 Postby **ColdFront77** » Fri Sep 17, 2004 11:16 pm

Putting the 120 hour forecast position close to your location is another issue. It is much easier for someone living in northeastern Ohio... like Floydbuster.

#19 Postby **MWatkins** » Fri Sep 17, 2004 11:23 pm

ColdFront77 wrote:Another issue is bringing the 120 hour position to your location. It is much easier for someone living in northeastern Ohio... like Floydbuster.

That assumes that people participating are people I disagree with in some sense.

So far i don't believe that to be the case with the 4 forecasts tonight. Also...as long as we don't get tracks from one paticluar region (all FL or SC for example) then the biases should balance out in the means.

If your idea is true...at worst we would see a left-of-track bias for Atlantic systems...however...we could eaisly do regression analysis to filter out blatent -removed- and exclude those tracks from the aggragate mean.

MW

ColdFront77 · #20 Postby **ColdFront77** » Fri Sep 17, 2004 11:25 pm

"I feel like I am a wishcaster because I like to experience interesting weather."

Of course I do nothing more than make a mountain out of a molehill. I would like to be one of the few to do this, but something is keeping me from posting my forecast.

OK...Time for a Storm2K Concensus Forecast...Please Join

OK...Time for a Storm2K Concensus Forecast...Please Join

MW Forecast for 10:30PMEDT 9/17/2004

10:30PM Storm2K Ensemble for Jeanne

Who is online