Template talk:Timeline Atlantic Hurricanes

I don't see much use to this timeline, and it's woefully deficient. Where's Ivan, for one thing? Are you only mentioning landfalls? Because Rita hit at 3, not 4. I don't really see anything that this adds to the main article, and it will just add a lot of empty space. --Golbez 13:23, 29 September 2005 (UTC)

ok ad, create and omit as much as you want. Ivan is only 3 and Rita is erased. The template should just be an ideo what can be done with timeline.

Well I just don't understand what the point of this is. To show all cat 4-5 storms that hit land? That's handled easier in text format, I don't really see how doing it in a timeline format helps much, but I could be wrong. :) If it's a test for you, perhaps you could do it in a personal sandbox. I like timeline, I've tried using it for single-season timelines, though it's not quite built for that (it doesn't like doing an axis by day). --Golbez 00:14, 30 September 2005 (UTC)

I agree, the timeline is not very useful and in many places downright wrong. (1935 is wrong, and most hurricanes list pressure at landfall but the most recent ones list lowest pressure)

Hurricanes and El Niño coherances

This timeline shows graphically in what time frame hurricanes concentrate. A simple matrix can't demonstrate it so well. It would be interesting, to show on the right side the timeline of El Niño in order to demonstrate the coherances with the Southern Oscillation Index (SOI).

No it doesn't, because (1) you have selectively chosen which hurricanes to put on the timeline and (2) hurricane data from before about 1965 is woefully incomplete and cannot be compared in the same way as modern data (if Rita/Katrina/Gilbert had happened back then they would not have known about their low pressure while out at sea). To get useful data over a long timeframe you have to restrict your comparison to only what is known from back then. For instance: between 1947-1965 there were 14 major hurricanes to hit the east coast of the USA, while between 1966-1984 there were 0. To find comparisons like this it is much more useful to look at published research (NOAA has lots of documents on this) rather than do your own original research. Jdorje 04:23, 6 October 2005 (UTC)

At User:Jdorje/Timeline is my own little pretty picture. Jdorje 05:41, 6 October 2005 (UTC)

That was my intention, to discuss the timeline hurricans. Thanks for using the Template. To create my chart, I got the data from http://www.nhc.noaa.gov/Deadliest_Costliest.shtml and selected only cat. 3 and above. --Saharadesertfox 23:17, 6 October 2005 (UTC)

If you have objective criteria for your list, you need to state what they are. You obviously didn't get it from the NHC site since that lists only USA hurricanes (it doesn't mention gilbert at all, for instance). You need objective criteria for choosing which hurricanes to include and what data to put about them. These criteria must be chosen in a way that does not bias in favor of certain periods of time - you cannot choose based on minimum pressure, since minimum pressure isn't known accurately for hurricanes before about 1980 and generally not at all for those before 1965 - you cannot choose based on max winds since these aren't known accurately for hurricanes before about 1995 and often not at all for those before 1965 and practically never for those before 1945 - you cannot choose based on damage cost, since this is highly random and (unless you choose the estimate-if-it-hit-today value) biased in favor of recent years - you cannot choose based on number of deaths, since this is similarly random and biased. My conclusion is the only thing you can look at is landfalling US hurricanes, since this is the only data known with any certainty (even then it's all estimates, but the NHC has lots of resources that they put to work estimating the hurricane's strength based on reports from landfall). Jdorje 23:39, 6 October 2005 (UTC)

Now, for what you're trying to show: you want to show a correlation with the ENSO. This is impossible (I think) since hurricane data is too random over individual years, while the ENSO is a year-by-year variation. My timeline shows approximately what the NHC claims: that hurricanes in the Atlantic (NOT the gulf or caribbean) vary greatly according to some decades-long north atlantic oscillation. Try to find the NHC article where I found the stat I quoted above ("between 1947-1965 there were 14 major hurricanes to hit the east coast of the USA, while between 1966-1984 there were 0"); unfortunately I lost it. They had all sorts of pretty diagrams where they showed some pretty startling things - such as that gulf hurricanes actually follow a pattern that's approximately the inverse of atlantic hurricanes. Remember that Wikipedia is not the place for original research, so what you REALLY want to do is copy the NHC's diagrams on this (which were quite convincing) and import them into the encyclopedia. Jdorje 23:39, 6 October 2005 (UTC)

Complete lack of criteria

This timeline lacks any criteria. The title says "hurricanes with highest S-S category" but it isn't. Why is the cat4 Charley included while the stronger cat4 Gloria and Floyd are not? Why is the minimal cat5 Isabel included while many stronger cat5 storms (Hattie, Dog, Beulah, 1947, etc.) are not? Why does Hugo, plus every storm from before 1965, have landfall pressure listed while other storms have their lowest pressure listed? I say it again: if this timeline is to be of any interest, it needs to have consistent and reliable criteria. Jdorje 23:52, 24 October 2005 (UTC)

The problem with the original format is that there is not accurate information for what you're trying to show. However if you look at something more specific - like landfalling US hurricanes - then the information you can find is accurate. The NHC has lists of every "major" (cat3+) hurricane to hit the united states, going back to either 1890 or 1851, and their data is very thoroughly researched. You can further (as I did in my timeline) separate these by location of landfall: to show the decadal pattern of landfalling east-coast storms. Jdorje 02:37, 2 December 2005 (UTC)

Horizontal timeline with complete new layout

I thought a while and tried to create a complete new timeline with all hurricanes with cats in 2005, the cats are in different colours. Unlimited hurricane seasons can be added. The small black lines in between the length of the hurricanes indicate the landfall. The timeframe can be changed from only July to October.

Here the result. Is it worth? --Saharadesertfox 00:12, 2 December 2005 (UTC)

That's kindof interesting. You should probably take it over to Talk:2005 Atlantic hurricane season. Jdorje 02:02, 2 December 2005 (UTC)

It should probably show all storms (including tropical storms, not just hurricanes). It might be good to have the colors delineate specific parts of the storm, so Katrina (for instance) would show red only for the portion where it was cat5 (two days), and other colors for the rest of the storm - but this might just make it look like a rainbow; I'm not sure. The extraneous months should be removed (the season runs June 1 - nov 30, though these bounds can be extended if the storms run over). Jdorje 02:35, 2 December 2005 (UTC)

I have altered the timeline with most of your suggestion. It is no problem to delineate the specific parts of the storm into the categories. But to get the data is not easy. --Saharadesertfox 18:34, 4 December 2005 (UTC)