I am using Network Analyst to run some analyses on global networks. In particular, I am trying to find shortest distances between points. However, I noticed that routes are not crossing the Pacific. There must be some disjoint at each edge of the network, rather than connecting from one side to the other. I assume the problem stems from representing a 3D problem in 2D.

As an example, a route from Vancouver to Japan will go south, through the Panama Canal, across the Atlantic, around the southern tip of Africa and then diagonal up to Japan. Of course, the shortest path is directly across the Pacific.

Any idea of how to remedy this issue?


As a workaround, you could connect the routes from -180° to 180° with "artificial" lines with an impedance of 0.

EDIT: another solution would consist in working with a network in a global azimuthal projection (e.g. Wiechel's) centered on the North pole. Then the discontinuity occurs in Antarctica, where you shouln't have too many routes.

  • Thank you for the suggestion. I understand setting the impedance, but how would you suggest I create/network the "artificial lines" in this case? – Drew Jan 5 '15 at 14:04
  • If there are not too many, you could draw them by hand (drawing the parallels that do not cross the dateline). – radouxju Jan 5 '15 at 14:50
  • Unfortunately that won't be a realistic option due to the number of lines required. Thank you, though. – Drew Jan 5 '15 at 20:05
  • I also considered changing the projection. I think that would work, but it just feels like there should be a more straightforward solution. – Drew Jan 5 '15 at 22:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.