Monday, August 20, 2012

Hurricanes Since 1851


Hurricanes and tropical storms since 1851, as far as we can tell.  Poster print options available here.  You are looking up at the Earth; Antarctica at center, the Americas to the right, Australia and Asia to the left, and Africa at the bottom.

What Is It?
Ok, here's a bottoms-up view of known tropical storms and hurricanes dating back to 1851.  The fine folks at NOAA keep an archive of storm paths with wind speed, storm name, date, among other attributes, and are always updating and refining information for past events based on historical evidence and educated hunches.  The data are awesome and they make it available in several formats.  Here's what it looks like slapped onto a polar projection (looking up at Antarctica) with point color tied to intensity...

A couple of things stood out to me about this data...

1) Structure.
Hurricanes clearly abhor the equator and fling themselves away from the warm waters of their birth as quickly as they can.  Paging Dr. Freud.
The void circling the image is the equator.  Hurricanes can never ever cross it.  Check out the oddball to hit South America, though.

2) Detection.
Detection has skyrocketed since satellite technology but mostly since we started logging storms in the eastern hemisphere.  Also the proportionality of storm severity looks to be getting more consistent year to year with the benefit of more data.

Anyways have at it, and you can find a superultramega-sized version here or check out poster print options here.  And check out the slightly animated version of hurricane seasons since 1978 here.

Projection
The best part about this process was finding a map projection that would be both cool looking and show the circuitous structure of the data.  I still prefer the South Pole Stereographic but here are some of the also-rans.  Maybe you'll find one of them more orienting...

Cool Mistake
Also when first registering the data, I accidentally assigned time as one of the location coordinates and, after the initial disappointment of what I though was mucked up data, realized that what I was actually seeing was a timeline visualization of where we've historically paid mind to collecting storm data.  Check out this bit of happenstance that showed me we only really started logging the East and South hemisphere versions of these things around 1978 (which is also the year that I ultimately started keeping track of things)...

Increasing satellite detection -particularly geostationary satellites, brought the southeast quad of the globe 'online.'  The two charts are a maybe-mind-bending look at the when and where of detection, showing the abrupt inclusion of hurricanes (known as Typhoons, locally) from other parts of the world in the data archive.

Other Sort-Of Related Maps
If this map is interesting to you, you might be amenable to plunging down the rabbit hole of related visualizations...

Tornado Tracks since 1950:
























36 comments:

  1. I can't get my head wrapped around what view of the globe this is from?

    ReplyDelete
    Replies
    1. I was confused as well, as you don't typically see pictures of the south pole like this. Antarctica is in the center, Australia to the left and the Americas to the right.

      Delete
    2. the center of the image is the south pole and it laid on it's side.

      Delete
  2. To get a perspective on this map. This map looks at the earth from the south pole. Antarctica is near the center. South America is to the right and Australia to the left. Africa is at the bottom. The earths equator is near the circumference of the circle.

    ReplyDelete
  3. You blew this too NOAA. Say no to corrupt NOAA!
    Do a search.." corrupt NOAA " you will see what I mean.

    ReplyDelete
    Replies
    1. That data from NOAA is really from the WMO, how do you like the UN?

      http://www.ncdc.noaa.gov/oa/ibtracs/index.php?name=wmo-data

      Delete
  4. This is a dumbass way to present the information.

    ReplyDelete
  5. Perhaps it was difficult for people living in the North Hemisphere, but being in Australia & New Zealand it was relatively easy for me once I actually could pick out the feint outlines of the continents — which was the hard part.

    Very cool though!

    ReplyDelete
  6. It IS rather difficult to figure out. If you flip your laptop or iPad so the right edge of the map is pointing up, you're in a north-south orientation for the Northern Hemisphere. If you're using a desktop you're on your own. Then, you can clearly make out S. America, and follow on up to Central Am. and N. Am., west edge of Africa. Australia is just below Antarctica, and "below" Oz is Indonesia, S. Asia. Flip screen to the left to get N-S orientation for S. Hemisphere.

    ReplyDelete
  7. Does anyone else feel odd about it looking very much like an eye?

    ReplyDelete
  8. This is a spectacular presentation of the data! I can't imagine it being as informative from any other perspective. Congratulations, and thank you so much for making a high resolution picture available so we could zoom in on specific areas.

    ReplyDelete
  9. What is the time interval between dots for each storm path? I didn't see it on the map legend.

    CJSF

    ReplyDelete
  10. What is the time interval between dots for each storm path? I didn't see it on the map legend.

    CJSF

    ReplyDelete
  11. Fabulous! Good way to show the data and turn it into information. Nice!

    ReplyDelete
  12. Awesome. Really, REALLY AWESOME.

    ReplyDelete
  13. I'd love to see this dissected into decades perhaps - to see when there have been more hurricanes and if they have shifted in their pattern of 'approach'.
    Great visualisation - unusual but makes more sense since it also displays the Coriolis Effect that causes hurricanes to curve. The high res. picture is awesome and terribly absorbing....
    Great stuff - thank you.

    ReplyDelete
  14. This is great, as I'm a hurricane researcher. But I'm wondering why in the West Pacific Ocean basin there appears to be a dearth of bright green (Category 5) dots? There should be plenty of storms that reach this threshold.

    ReplyDelete
    Replies
    1. I think this is just an artifact from the fact that the data have only been collected since 1978. Category 4 hurricanes happen often enough that they saturate the map in fewer than 30 years, but Category 5 are rare enough that it would take some longer amount of time to saturate the map. I think the map doesn't really display density of hurricanes, just whether there has ever been one relatively nearby of a certain category.

      Delete
    2. Hmm you may be right, though I'm still surprised that even going back to 1978 produces what appears to be only a handful of cat 5 storms. If that's the case though, that's too bad because it gives the impression that the Atlantic is much more active (and with stronger storms) than the West Pacific, which is not at all the case.

      Delete
    3. I see quite a few around northern Australia and Bangladesh. Keep in mind that before satellite tracking a lot of storms in the Pacific were poorly tracked, I assume that's why data here started in 1978.

      Also, keep in mind that globally we're at a 30 year low in storm activity. The North Atlantic has been busy (like the 1930s-1950s) thanks to the positive AMO. The Atlantic is also a lot smaller than the Pacific, I think density-wise it does have more storms.

      Delete
  15. wow, the structure of both pictures is fascinating. Both the circular and linear. Thanks!

    ReplyDelete
  16. I can see that if you are American, and America is not in the centre of the map, you would be confused...

    ReplyDelete
  17. This comment has been removed by the author.

    ReplyDelete
  18. This one is from NOAA and a lot easier to understand, but I don't think it goes back that far.

    Click the Hurricane radio button once on the site.
    Historical Hurricane Tracks
    http://www.csc.noaa.gov/digitalcoast/tools/hurricanes

    ReplyDelete
  19. Congratulations for getting this posted as the Astronomy Photo of the Day at http://apod.nasa.gov/apod/astropix.html - "Each day a different image or photograph of our fascinating universe is featured, along with a brief explanation written by a professional astronomer."

    That's a high standard to uphold, the post has been dissected at http://wattsupwiththat.com/2012/09/04/nasas-astronomy-picture-of-the-day-double-blunder/

    Cool images though. I'm pleased you even have the storm that formed off the Brazilian coast several years ago.

    ReplyDelete
  20. Impressive work M. Nelson!

    The cute kid on your back will be proud of his/her father very soon.

    inSitu Communications

    ReplyDelete
  21. More on that Brazilian storm, Hurricane Catarina, http://en.wikipedia.org/wiki/Hurricane_Catarina .

    Your track is much shorter than the track on the Wikipedia page, or the (apparently) much more complete image at http://en.wikipedia.org/wiki/File:Global_tropical_cyclone_tracks-edit2.jpg

    Does the WMO data include tropical storms or just hurricanes? Your text is in unclear - sometimes you say "hurricanes" sometimes you say "hurricanes and tropical storms."

    ReplyDelete
  22. I'm wondering what software you use for all your mapping visualisations. This isn't something you can just do in ArcGIS or QGIS as far as I can tell! It looks more like something custom done with Processing...

    ReplyDelete
    Replies
    1. Hi Barry, thanks for the comment. GIS software will only get you so far, typically. It's an analysis tool that isn't ever going to have the same visual editorial horsepower as a design package like photoshop or what have you. Fair enough. Specialization drove the industrial revolution. I happened to use Adobe Fireworks for the post-processing but my visual style is lazy and doesn't require a lot of effort.

      Delete
    2. Great thanks. I was fishing to see if you were using Open Source components for possible inclusion in an Open Source Atlas project I'm working on! I've just been looking at the Fireworks web page and I'm still not sure what it does for you... Anyway. Nice.

      Delete
  23. Your image is glaringly incorrect at first glance to anyone familiar with tropical cyclones.

    You should understand and interpret the data that you use, and research to make sure that you have done so correctly. Not all RSMCs use 1 min wind averages, most use 10 minute winds. The Category 1-5 designation is only valid in the EPAC and ATL basins, but can be interpreted to the other basins. NASA has a correct image. Check it out.

    ReplyDelete
    Replies
    1. 10 minute averaged windspeeds are lower than 1 minute averaged windspeeds. Also it is a little tricky to apply the Cat 1-5 intensity to data in the other basins. One thing you could do is ask NASA for their raw data for their image, or their methodology, and regenerate your map based on that data.

      Delete
    2. Margie you might be better off voicing your concerns to the folks at the Hurricane Best Tracks Archive, if you think the data is off. This is a direct mapping of their wind speed calculations. There comes a practical point where in order to make any map I have to trust the data of a vetted provider (in this case, the National Climatic Data Center of NOAA).

      Delete
  24. I would love to see a similar map with a polar viewpoint of the Sun used to project Sunspot data.

    ReplyDelete