I have a large hydrology geometric network with more than 300 sinks and roughly 150 barriers. I'd like to have the SinkID added to the attributes of the barriers (some individual networks have a single barrier, some have multiple) for later analysis but haven't seen any obvious way of doing so. I could assign the SinkID to the first stream segment to which it's attached but some barriers may be 10 or more stream segments upstream.

enter image description here


  • It' hard to make sense of the map without a legend, but I suppose the "barriers" are watershed boundaries (shown in gold), right? But what are the sinks and precisely what are their relationships to the barriers? – whuber Jan 10 '12 at 17:07
  • Sorry that wasn't clear @whuber. The Sinks (green points) are the outflow points for streams and rivers. Barriers (red triangles) are points of obstruction (culverts, etc.). The gold boundaries are just watershed boundaries. – phyllo Jan 10 '12 at 17:39
  • Is a programmatic solution acceptable? – Kirk Kuykendall Jan 10 '12 at 18:22
  • Because I was wrong, let me make sure the situation is clear: "barriers" and "sinks" are both locations along streams. Associated with each barrier will be at most one "sink"; namely, the one (if any) first encountered when moving downgradient from the barrier. This indicates that your problem is standard (with lots of available solutions): if you make these sinks the "pour points" of a watershed analysis, you simply want to associate each barrier with whatever watershed it lies in. – whuber Jan 10 '12 at 18:28
  • There are in some cases multiple barriers associated with one sink (a large river network with numerous road crossings). The watershed boundaries in some cases do have multiple sinks as this is a coastal area and there may be many small streams emptying into the ocean in one delineated watershed. I could divide the watersheds up so that there's only one sink (pour point) per watershed and then associate the barriers with watersheds, as you suggested. I probably should have split the watersheds earlier. – phyllo Jan 10 '12 at 18:43