How to derive reach breaks

[MichaelStead]

I was recently asked to create maps of mainstem stream slopes within a bunch of watersheds classified into 5 slope categories. I was also asked to find changes in slope greater than 2%.

I have done this a few ways, but due to the need to find changes in slope I decided to do this as a routed network so I could manipulate external tables in Excel. I have a 10m DEM so I split my line into ~ 10m sections and converted to points. I used values from the DEM, interpolated for the location of the points to populate this point feature with elevations, and then I used the locate along line tool to create a route event table. I then manipulated this table in excel to be a line event table by making each points F_MEAS the preceding points T_MEAS and then using this to calculate a slope and a change in slope field. From this I can create a line route event layer for segment slope, and a point layer for change in slope.

The person I did this for finds this level of detail the equivalent of noise so I suggested that he define some reaches and I could average the slope. In response I was sent the TFW Stream Segment Identification manual with a methodology involving 3 mylar overlays and no GIS. I subsequently found this http://www.trentu.ca/iws/documents/StreamSegmentationandCharacterizationReport2006.pdf methology that uses ArcHydro.

Does anyone out there have any methodology or tools they can point me towards? I have only defined reach breaks in the field.

Thanks,
Mike

[DuncanHornby]

Mike,

Your technique is how I create reaches along rivers, so you are not the only guy out there struggling!  In the past I've tended to scale the reach by something like strahler order or catchment size. This method will generate overlapping reaches.

Your existing 10m long line events will have the route ID and a start/end measure. You could try to start with the line nearest to the source and track downstream and identify when there is a big change in slope and identify that as your reach break? Thus lots of small changes in slow would be aggregated into a single reach. I'm not sure how this will work at tributary junctions at depends how you built your route layer.

This will certainly require some coding by you.

Duncan