ArcHydro 10.7.1: Stream Segmentation

[EmmaAppel]

I am trying to use the stream segmentation tool to delineate catchments within a watershed in order to *hopefully* identify wetlands that have no upstream contributing wetlands. However, when i run the stream segmentation tool i am getting an unexpected output that looks like a grid of squares. After running the catchment tool this is confirmed and the catchment areas are just squares that actually cut most of the streams off in unexpected places. 

It is my understanding that the stream segmentation tool identifies streams as headwater streams, links, and output streams so the boundaries should follow individual catchments. Is this understanding correct? I have no idea what i am doing wrong, as my input flow direction and stream grids were both derived from the same enforced DEM. I have attached the result of my catchment analysis to illustrate my issue. 

any help would be greatly appreciated. Thank you. 

[MarkBoucher]

Emma,

My first suspicion is that the inputs to the Stream Segmentation tool are not correct. They should be (at a minimum)  the flow direction grid the stream grid.

Mark

[EmmaAppel]

Mark, Thank you for the reply. 

The inputs that I am using are a stream grid and an enhanced flow direction grid. From the User Guide, I understand that those are the only mandatory parameters. can you confirm/ clarify?

[MarkBoucher]

I have not used the "enhanced flow direction grid". But I believe the stream grid (str) and a flow direction grid are all that are needed to create the stream link (strlnk) grid. The str and strlnk look exactly the same except the str would have 1's and 0's but the strlnk grid has 0's (where no stream is indicated by the flow accumulation grid threshold for str) and integers of 1 and greater.

That is what's show on page 37 (pdf pg 38). in the document at  http://downloads.esri.com/archydro/archydro/Tutorial/Doc/Arc%20Hydro%20GP%20Tools%202.0%20-%20Tutori....

It would be interesting to see what your str grid looks like.  Looking at the image you shared, it makes me wonder if your rawdem is right.