I want to ask you what method do you suggest me for to trace stream network in flat areas and please explain me how to deal with the previous preparation of the DEM.

  • 2
    That's a bit of a broad question. Have you tried anything? – John Powell Nov 28 '14 at 19:07

The whole idea of DEM-based flow path modelling is that topography is the main controlling factor for determining where overland and near surface water will travel and that a DEM can be used to derive that topographic information. However, when you have a flat area, that assumption breaks down significantly. In flatter regions topography, at least at the scale of a relatively coarse-resolution DEM, is no longer the controlling factor for water redistribution. Instead evapotranspiration, soil characteristics, properties of the regolith, vegetation distribution, micro-topography, and many more factors become important for determining the pathway of streams. As such, all algorithms for hydrologically 'correcting' a DEM to enforce flow along a flat area are simply enforcing an arbitrary solution and not one that we should expect to be related to the physical reality of the stream network. Just about the only thing that you can do to improve the situation in these flat areas is:

  1. get a finer-resolution DEM with an improved vertical precision, such that you can start to account for the impacts of finer-scale topographic variability (e.g. LiDAR data), or
  2. burn in a mapped stream network.

Now, I'm really not particularly fond of stream burning in general, but when you have a flat area it's likely better than most of the flow enforcement methods. The other issue you must consider is why do you have the flat area? Is it because it is really an area of low relief in reality or is it the result of a lower-than-needed vertical precision DEM (applying a rounded-to-nearest-metre DEM to hydrological modelling in the Prairies may not be appropriate), or is it because you've added artefact flat areas to the DEM by performing depression filling. If the later is the case, you should consider using depression breaching rather than filling to help solve this problem. Good luck and thank you for asking this interesting question.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.