How to create a multimodal network dataset that includes land routes and sea shipping routes

Honki

I want to generate the shortest transportation route from a certain city in Country A to a certain city in Country B. Therefore, a network dataset containing land roads and sea routes needs to be created.

At present, I have the datasets of global land roads and sea routes, but they are independent of each other and have no intersection points. I also have a dataset of global ports. May I ask how to create a network dataset that can solve the shortest path using Network Analyst module?

In addition, in order to make the solution of the global network routing possible, how should the spatial reference coordinates be selected? I have currently attempted to use the coordinate system of WGS 1984 to solve the shortest route of global shipping routes and found that the shortest path cannot cross the Pacific Ocean. See the following picture, the grey lines are the roads and shipping lanes.

 

If you have any good solutions or suggestions, please tell me. Thank you very much.

RodneyConger

Hi Honki, This is and ambitious endeavor. Seems like you are off to a good start. The coordinate system you'll need to use for this is a Polar projection. I've had the most luck with North Pole Azimuthal Equidistant. That one will allow you to route across the antimeridian. You can connect the land network with the sea network via ports by putting both classes in the same network dataset and supplying virtual edges representing connection. The virtual edges could be a third. You'll need to know something about the mode transfer if you want to model time or monetary cost. If you are just solving for distance, that's easier. If you want to make this network even more realistic, you could look into shipping lanes derived from AIS data.

My team at Esri has quite a bit of experience helping customers build these networks. Let me know if we can help.