AnsweredAssumed Answered

How delineate sub-basins for urban storm drain network?

Question asked by AMRooke on Feb 16, 2016
Latest reply on May 1, 2018 by dalmond

I am struggling to find a way to delineate sub-basins for an urban storm drain network, with raster terrain and vector outlet points (hundreds of storm drain inlets).  I had hoped to burn in the storm drain inlet locations with lower elevations in the terrain, to force sinks at those locations, then I could just delineate basins using small stream definitions, then aggregate the mini-sub-basins together.  This would be a bit of work, but I can't even get this to work, as when I take the inlets point shapefile (with a "-10" value for all records in the elevation field) and create a raster, it only creates at the discrete inlet location points, and when I add it to the basin terrain raster, it only outputs the revised elevations at the inlet locations, as it seems that every raster math I have tried will not create a value if there is NoData at one of the input rasters.


Ideally, there would be a way to create a basin with multiple outlet points (I need about 150 for my current model), but it appears that neither ArcHydro or GeoHMS have the ability to work with more than 1 outlet.  Trying to generate the sub-basins individually would be a nightmare, as you would have to start at the uppermost inlet, then to the next downstream inlet and manually remove the area of all upstream inlet capture areas, and repeat until all areas had been identified.  Unless some automated way is available, I will end up just manually delineating the areas by following contours, as has been done for ages.


I have use ArcHydro/GeoHMS for delineation of surface hydrology (i.e. without stormdrain inlets and underground conveyance) for 20 years, and that's a piece of cake, but I have been beating my head against a wall trying to solve the urban drainage problem.  If somebody knows of a helpful tool, I would appreciate learning about it.  If there is not one yet, it does not seem like it would be a difficult exercise for one to be created by a person with some GIS coding skills (perhaps I can get one of David Maidment's students to do this).