This forum is shutting down! Please post new discussions at

Delineate riparian zone (buffer)


I'm trying to delineate the riparian zone (buffer of 30 m and 60 m; DEM 30 m resolution). But the SDR model's output raster stream.tif does not exactly overlap with my observed stream file. Could it be possible to delineate the riparian zone buffer from the Invest stream raster? What other solutions / tips do you have on this issue?



  • RichRich Administrator, NatCap Staff
    You could try a hack where you crank the threshold flow accumulation parameter so high that no streams are classified, then plug in your own stream layer in the optional "drainages" parameter.  Realize in the end though, the model still routes flow based off the DEM, so I'd expect some error if your stream layer doesn't align with the landscape drainage, but maybe that's good enough for what you need?
  • Hi Rich

    I followed your recommendations. At the usle.tif and the sed_export.tif,  the value inside the dranaige pixel is zero (OK). However, at the sed_retention raster, the value inside the dranaige pixel (in stream_and_dranaige.tif) is diferent of zero (not expected) , what does not happen with the pixel of Stream.tif raster - output of the model - (the value is zero in the pixel that correspond the stream). 
    So, I don't Know if I could use the  stream_and_dranaige.tif to generate my riparian buffer, given this pixel's value for sed_retention.tif.....
  • RichRich Administrator, NatCap Staff
    Oh interesting.  That drainage layer is a hack we didn't think through too much when we did it back in the day.  I'm not sure what to recommend here.  I'd need to consult w/ Perrine before we made a change to the model like this, and I'm afraid the change might be to remove the option of a drainage layer entirely.  But maybe we should zero it out... 

    Perrine is travelling at the moment and NatCap's only hydrological authority.  I'll get back to you when I can! 
  • Ok Rich!

    Thank you
  • PerrinePerrine Moderator, NatCap Staff
    Hi there, 

    Quick note to say that I don't believe the erroneous sed_retention values are a big issue. I agree with Rich that for consistency, the model should probably have sed_retention=0 on drainage layer pixels (since these pixels are considered as stream pixels).
    But remember that sed_retention is an index value: see definition in the User's guide here. It is NOT the amount of sediment coming from upstream and deposited on a pixel, so make sure that you use this index in an appropriate way. Since all the "drainage layer" pixels are stream pixels by definition, the general assumption (in this model) is that there's no retention there so the index value doesn't matter (but could be set to 0).

    Finally: I agree with Rich that the model will use the DEM's hydrologic routing. Therefore, it may be preferable to use the stream.tif layer created by the model to compute your riparian buffers --so that the flow paths actually converge at the lowest points of hillslopes. Having said that, I also understand that you may have other reasons to use your own stream layer --so it's a decision you need to make depending on your modeling objectives.

    Hope this helps!



  • Hello

    Thank you for the answers, very heplfull.

    But I still have questions about sediment retention.

    As model's output there are:

    • sed_retention.tif (tons/pixel): Map of sediment retention with reference to a bare watershed.

    • sed_retention_index.tif (tons/pixel): Index of sediment retention, used to identified areas contributing more to retention with reference to a bare watershed. This is NOT the sediment retained on each pixel (see Section on the index in “Evaluating Sediment Retention Services” above).

    • watershed_results_sdr.shp  >>>   sed_retention (tons/watershed): Difference in the amount of sediment delivered by the current watershed and a hypothetical watershed where all land use types have been cleared to bare soil.

    a) sed_retention.tif (tons/pixel): how this is calculeted?

    b) sed_retention_index.tif (tons/pixel): This one is  is computed by the model as follows, right?

    c) watershed_results_sdr.shp  >>>   sed_retention (tons/watershed): For one is  the model uses as a benchmark a hypothetical scenario where all land is cleared to bare soil: the value of the retention service is then based on the difference between the sediment export from this bare soil (i.e Sed_Export_Bare Soil = Sum of all pixels of [rkls.tif (tons/pixel) X intermediate_outputs / sdr_factor_bare_soil])   and  the sediment export from scenario of interest (i.e sed_export (tons/watershed)table).
     Is that right?? 
Sign In or Register to comment.