I'm trying to run an analysis where I need to clip a radiating lines layer by land features to get the maximum distance from the central point to the nearest obstacle. What I have is a large line layer containing approximately 40,000 points, each with 32 radiating lines extending out at equal intervals (e.g. 11.25 degrees) and a vector polygon file representing coastlines and reefs.

The problem I'm having is that the "Clip" and "Erase" functions remove the sections of line on land but when a line encounters a small obstacle the tool only removes the section over land and leaves the portion of line which continues on afterwards (e.g. adjacent bays, other side of peninsulas, etc). This is especially problematic in regions with complex coastal features.

What I need to do is remove the hanging lines after the nearest land.

I've looked into line of sight analyses but the only one's I can find rely on 3D surfaces. Does anyone know of a 2D version where all line fragments will be removed after reaching the nearest obstacle?

This needs to be as automated as possible as I'm dealing with large datasets (24 to be exact, each with 40,000 points).

The end goal is to produce a layer to calculate wave fetch.

1 Answer 1


if you have a feature at the start of your radiating lines, presumably, all the line features you want are touching this point. so after clipping the features based on the water polygon (and possibly exploding [advanced editing/explode multipart feature tool]), you could use a select by location to only select the lines that touch a center point, then invert your selection, and delete the excess line features.

  • Thanks, but I will be dealing with >900,000 points, the radiating lines from each point intersect those of nearby points, and they are not multipart features. So I can't explode multipart features as there are none. I also can't use trim line because of the intersections between lines. I did find a model that will extend my lines to the nearest feature and this may be the way to go instead of trimming down the lines. However, the way the model is setup it trims the lines to the first feature they encounter (sometimes another radiating line) so I will need to alter the model somehow. Commented Jan 22, 2014 at 15:44
  • One possibility to work around the overlapping features would be to convert the origin point of the radiating lines into a point layer, and try the select by location against that, which would have a pretty low chance of being intersected by other lines, and since you have a set number of radiating lines per feature, a basic summary count would quickly tell you if you are missing anything
    – TDavis
    Commented Jan 22, 2014 at 16:00
  • True, but my points are laid out in a grid (~1km cell size) and the radiating lines extend out up to 50 km from each bearing meaning they will intersect with each other, unless I'm missing something you're saying. Commented Jan 22, 2014 at 16:19
  • I wish I was that lucky. I think you're right and a model is the way to go. I've been working on something but it needs modification to run correctly. I've created a new thread about it here (gis.stackexchange.com/questions/83882/…) Commented Jan 22, 2014 at 16:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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