Stream vector for DEM reconditioning

Question

Hi,

I have a stream feature class I would like to burn in using DEM reconditioning.

what I’m wondering is: are there any specific requirements for the feature class?

I know all streams should either exit the DEM or end up in sinks but is there anything else to think about that is really key? I read that pseudo nodes should be removed and the geometry checked.

I am only really interested in getting drawing my catchments (not planning to do any complex analysis beyond this really).

Any suggestions are really appreciated.

Thanks a lot!

Alex

AlexanderVentura · Answer

Hi Dan,

Thanks for your answer.

Actually I’m not getting unexpected result, just want to make sure I’m following “best practice” or at least not making serious mistakes. But if all streams get burnt in and have the values I expect it should be fine right?

My stream feature class is actually 2 separate ones that I merged into one and then use to burn in streams. The reason is I believe the 2 separate stream feature classes are for main and secondary streams. They are generated by the same national authority and in order to include all of the streams in my DEM reconditioning I merged them.

Also I am not clipping the streams to the raster extent as this way all my streams “exit” the DEM and won’t get refilled after burning in. Is that a problem and is it better to instead clip the stream feature class to raster extent and then extend the streams beyond the DEM manually (by editing streams)?

One last question if you don’t mind: my DEM raster has values (approx 0 meters) for the ocean. As many of my streams end up in the ocean I was thinking of converting the ocean polygon to raster and removing this from my DEM. This way all streams will exit the raster (and won’t get refill after burning) and I can add “coastal” catchments at the end of the terrain preprocessing to complete areas with missing catchments. This makes more sense to me than treating the ocean as a “sink”.

Does this make sense to you?

Thanks again for the help.

Alex.

DanPatterson_Retired · Answer

are you getting unexpected results?

I would convert the stream to a raster with the same extent and cell size first using the dem as the snap raster.

In that way, you have control over the result and are sure that the conversion of the stream to a raster is done overtly and not behind the scenes where you can't inspect it