Sunday, November 24, 2013

The longest straight line you can sail on Earth ?

The longest straight line only touching water ?
Note : the line does not appear straight on this image due to the Plate Carrée 2D projection. 
see Great Circle demo
 - map from gcmap -

source : wikipedia
From the south coast of Balochistan province somewhere near Port of Karachi, Pakistan (25°25′N 66°25′E) across the Arabian Sea, south-west through Indian Ocean, near Comoros, passing Namaete Canyon, near the South Africa coastline, across the South Atlantic Ocean, then west across Cape Horn, then north-west across the Pacific Ocean, near Easter Island, passing the antipodal point, near Amlia island, through the South Bering Sea and ending somewhere on the east-north coast of Kamchatka, near Ossora (59°38′N 163°24′E).
This route is almost 32,000 km (20,000 mi) long.

Actually, the video shows the best representation of “the way things are”:

kml file for use with Google Earth (17302 Nm / 32043 km)
GE “ruler” tool allowing you to calculate “straight line” distances over the globe
from Kamchatka Peninsula in Eastern Russia (south of the end of the Aleutian Islands archipelago)
to a point near to Graham’s Land (the long finger-like peninsula on Antarctica that points toward the Falkland Islands),
then directly between Madagascar and the African continent for ending in Pakistan

The longest straight line only touching water ? Really ?

Who would have guessed you could sail in a straight line from the Kamchatka peninsula in Russia to Pakistan ?
Actually not really because the straight line crosses over the Aldabra islands in the Indian Ocean...

 Aldabra atoll

 crossing line about 15 km inside the Alabra atoll in the North,
but only 7 km outside Anjouan island (Comoros) in the South

Note : by the way, if you tried to sail it you’d probably be sunk by drifting icebergs in the Southern latitudes, but that’s beside the point.

Southernmost passage (latitude : about 61°30' S)


  1. I posted that original solution in 2006, but someone else made the Youtube video and probably just put up a "close enough" line.
    Here's my Facebook post on the subject...

    About 10 years ago I posted a challenge on the Google Earth (GE) forum to
    find the longest straight line you could travel over water before hitting land again.!topic/gec-fun-games/WZ3P_IvO0pc
    That forum is now archived and no one can post on that thread. But it's still there for all to see.
    I posted under the username Kryten (-after my favorite mechanoid, my standard name on all forums).
    My solution was pretty good, and it doesn't appear that anyone could find a longer path (-my solution was 19820 miles).
    Years later I saw it referrenced in a UK article and Ken Jennings posted something on it. But nobody ever linked back to that forum.

    But now, using Google Earth Pro, I was able to modify it with two great circles that split the
    globe, each with a radius of 6225.25 miles (with centers from opposite points on the earth) and have verified the length
    with these new coords as at least 19914 miles-- nearly 100 miles longer!

    Here are the endpoints.
    Karachi end:
    25d 20' 41" N
    66d 33' 55" E

    Kamchatka end:
    59d 45' 00" N
    163d 22' 25" E

    Some points in between:
    54d 10' 20.62" S
    3d 4' 12.91" E
    This is about 17 miles NW of Bouvet Island

    28d 21' 5.77" S
    111d 23' 17.18" W
    This is about 143 miles SW of Easter Island

    You can recreate my line using those 4 coordinates.
    It's very difficult to get a straight line that long in GE. You have to use multiple lines or a path with multiple segments.
    My proven solution uses a circle that splits the earth evenly, making it a "great circle". (Anything that isn't a great
    circle isn't really a straight line-- you'd have to constantly veer to the left or right by some amount.)

    1. " two great circles that split the
      globe, each with a radius of 6225.25 miles"

      If a circle's radius on Earth is 6225 miles, it's not a great circle.

      Guess the main problem is people are trying to get Google to draw the lines for them, and it won't cooperate. Better plan: get the latitude-longitudes from Google and do the calculations yourself. I'll check your four points above -- suspect there are slight turns at the ocean points.

    2. Turns out if we assume a spherical Earth, the great circle from 25-20-41N 66-33-55E to 59-45N 163-22-25E passes latitude 54-10-20.62 S southbound at lon 3-02-48.29 E; it passes lat 28-21-05.77 S northbound at lon 111-18-17.54W. I haven't checked whether that great circle misses all the islands.

  2. Hey, I'm the creator of that video. Mike Sutton, the discoverer of the line, is spot on, I screwed up when I made that video. I made it quickly simply to demonstrate how a 2D projection of a great circle was indeed a straight line. I did not spend more than five minutes on it, and it is certainly not accurate. The real line does not hit land when you map it accurately using the Great Circle Mapper.

  3. A probably-legit track on the WGS84 spheroid -- 32000+ km (it misses Aldabra, and everything else AFAIK)

  4. The great circle from 25-25N 66-25E to 59-38N 163-24E does hit Aldabra, but of course that doesn't "debunk" Pakistan to Kamchatka -- plenty of other possible starting/ending points that miss Aldabra, and some of them miss all the other islands too (far as we know).