Hi All,
I am trying to delineate multiple urban watersheds in GIS for the City of Denver (sometimes called sewersheds) using Spatial Analyst. I am looking at several drainage points along the South Platte River and trying to determine their catchment areas to understand what land uses/areas are draining to each drainage point. These points include surface water gulches and underground outfalls.
Up to this point, I have a relatively high resolution DEM and effectively 'burned' the storm sewer mains into the DEM to account for city drainage and get reasonable drainage paths for the underground outfalls. I was getting some weird looking watersheds for the surface water gulches as their flow accumulation paths weren't showing up. As a result I burned in the surface gulches and the South Platte River as well.
Half of the sites are on the left (western) side of the Platte and the other half are on the right (eastern) side of the Platte. The watersheds on the left side of the Platte look good as drainage goes from left to right into the Platte River. However, the watersheds on the right side don't work, as I've noticed that the flow direction is going the wrong way on GIS. Theoretically, it should be going right to left (east to west) draining into the Platte River; however, the flow direction raster GIS computed has the flow direction going left to right (west to east) which is wrong. It doesn't make sense as the elevation is increasing along the flow paths west to east, so GIS is saying that flow is going uphill. Technically the burned gulches and storm mains are at the same elevation, so maybe this is the problem? I've filled the DEM too. This thread had the same issue but I wasn't able to pull any solutions from it: https://gis.stackexchange.com/questions/108387/flow-direction-generated-for-river-is-incorrect
I'm not sure what to do now, so any advice is greatly appreciated! Also, sorry this is only my second question I've posted on ESRI, so apologies if there is a better place to post questions. Thanks!
Solved! Go to Solution.
Thanks for the tip, Mark. I actually was able to fix the problem by playing around with the burned streams some more. Storm mains weren't connected to the River, so I added polylines to connect the storm mains and the river and then reburned the DEM. This did the trick and GIS was able to determine the correct flow direction.
The flow direction is based on a filled DEM. If the filled DEM fills incorrectly, you can get flows going the wrong way. You can also train the flow direction if the results are mostly right, but the direction in the burned streams is incorrect. Maybe you need an outer boundary on the right side.
Thanks for the tip, Mark. I actually was able to fix the problem by playing around with the burned streams some more. Storm mains weren't connected to the River, so I added polylines to connect the storm mains and the river and then reburned the DEM. This did the trick and GIS was able to determine the correct flow direction.