Streams ending into the sea

454
3
02-26-2019 12:22 AM
Labels (1)
AlexanderVentura
New Contributor II

Hi all,

I was wondering if anyone could confirm the validity of my approach. 

I am running a terrain preprocessing using a raw DEM and a feature class for known streams (that I will burn in through DEM reconditioning). 

My DEM raster has values (approx 0 meters) for locations corresponding to the sea. Most of my streams (in my stream feature class) "exit" the DEM, however several individual streams end up in the sea. I was thinking of converting a sea polygon to raster and removing this from my raw DEM. This way I am making sure all streams will exit the raster (and won’t get refilled after burning). Then I can append “coastal” catchments at the end of the terrain preprocessing which will add missing catchments not associated to modeled streams. 

This makes more sense to me than treating the sea as a “sink" (and basically creating a sink structure for the sea). Does anyone have any comments regarding this approach and does it sound reasonable?

Thanks so much for the help.

Alex.

 

0 Kudos
3 Replies
DanPatterson_Retired
MVP Emeritus

burn the sea as well so that its elevation is the same as the burn depth... then it should end up back at zero.  Otherwise your approach seems reasonable

0 Kudos
AlexanderVentura
New Contributor II

Hi Dan,

thanks but how do I “burn” the sea? Should I use the “Create Sink Structures”? But this way the streams end up into a sink no? So essentially I would be treating the sea as a sink  and would need to add a sink point at the end of each stream so that it doesn’t get refilled. Or am I missing something?

0 Kudos
DanPatterson_Retired
MVP Emeritus

I can only think of the spatial analysis to make the sea a sink using the Con statement

Con("your_dem" == 0, -2, "your_dem")  # assuming a burn depth of 2 meters.

Arc Hydro must do things differently, I usually burn after sinks are filled to provide a preferential path for streams generate.  so go with your approach and test it, if my suggestion confuses the issue

0 Kudos