1

I've been playing with arcHydro and related processes a bit and have run into an obstacle. I have been creating stream networks from the flow accumulation, direction, and stream order functions. When converted from raster to polygon each segment of the streams is a unique feature within the output feature class. I would like it so that if there is no change in order from one segment to the next, that it is the same feature. Easy enough, dissolve.

The issue now is that where two streams of the same order come together to form a stream of a greater order, the two smaller order streams dissolve into 'one' stream as they share a vertex (their shared outlet which is the inlet of the greater order stream subsequent). Difficult to describe without visuals but I'm hoping this was clear.

EDIT I'm attaching some visuals to help clarify.

enter image description here

The first image is the layer as it converts from raster; with each line segment being it's own feature.

enter image description here

The second image is following a dissolve. The 'conflict' areas are those circled in red, where a common vertice has caused two streams of the same order to dissolve together when in fact they only share an outlet. For my purposes, I don't want this to be the case. Hoping the visuals help clarify.

Improve this question
11
  • 1
    Could you provide more info on the raster with the streams. If there is a table associated with the layer could you please provide or describe the table.
    – Reisenrich
    Jul 31, 2017 at 18:36
  • Yes of course. I have filled a DEM, with the filled DEM a (flow) directional raster was created. From the Directional raster a (flow) Accumulation raster was created, and from that accumulation raster I established a minimum accumulation value and reclassified the accumulation raster based on whether or not it is a stream (whether or not it meets the minimum accumulation criterion. The Stream Order process created a raster which held values indicating the order of a given cell (given it meets the accumulation threshold).
    – UnsoughtNine
    Jul 31, 2017 at 19:09
  • This raster was converted to a polygon feature class and the order field carried over to the attribute table - I reiterate stream order is stored within a field in the attribute table. Any better? I can try to gather some visuals...
    – UnsoughtNine
    Jul 31, 2017 at 19:10
  • Dissolve with no attributes to dissolve. Transfer order later. Why polygons? There is a tool raster to stream
    – FelixIP
    Jul 31, 2017 at 19:16
  • Dissolving with no attributes would only expand the issue as it would make the entire network one feature. I'm looking to have segments of the same stream (and same order) the same feature. I can't seem to find the raster to stream tool?
    – UnsoughtNine
    Jul 31, 2017 at 19:29

1 Answer 1

Reset to default
1

I have come up with an approach:

  1. Create a point layer with points for each vertex in the stream network layer
  2. Buffer each vertex a (small) desired amount (I arbitrarily chose 10 metres)
  3. Run intersect and create points as the output, this will create a series of points where the buffer intersects the stream layer at the setback amount.
  4. Identify which points of intersection are of the desired order or greater. Remove those that are not.
  5. Identify the smallest remaining order within the buffer. Any points greater than the smallest order (i.e the stream being flowed into) will be removed.
  6. Left with the smallest order points which are greater than the specified minimum order of concern.
  7. These are the pour points (automated). I snapped the pour points back half of the original setback (5m). So the pour point should be 5 or so metres from the actual outlet (dependent on the shape of the stream), which makes for clean delineation.

Hoping the brief description is clear.

Improve this answer

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.