I am currently working on constructing a countywide riparian area, which normally consists of the integration of water bodies, wetlands, wildlife corridors, flood prone areas and hydric soils. We perform such analysis before but this time around we would like to add the component of Steep slopes (15% or more) that are directly adjacent to riparian features that associated with a waterway. In other words, steep slopes that are directly adjacent to waterways or directly adjacent to any other riparian feature associated with the waterway will be included.

Figuring and extracting steep slopes is straight forward, the challenge remain on the proper methodology and tool to determine the slopes that flow directly down to other riparian components of a waterway. I cannot just extract the steep slope and merge wherever it intersect with the existing riparian area, since many of them do not facing the river. I need to figure out some steps to “draw” a line.

I am aware that some Hydrology tools from the Spatial Analysis can be used to define Watershed (such as using the function Fill, Fill Direction, Flow Accumulation, Snap Pour Point etc from the DEM file) I am not sure (a) if these procedures are applicable to what I am looking for, and (b) I wonder if is that a necessity to carry out all these procedure in order to get a particular section of the slopes that happens to flow downhill to riparian area.

Do anyone of you have any takes or suggestions on these?

Update: I have done the following so far:

  1. Construct the Riparian based on Wetland, Wildlife Corridor, Floodplains, Alluvial Soils & Water Bodies.

  2. Using the Spatial Analysis on the DEM to find out the slope.

  3. Using relcass to find out the steep slopes that are more than 15%. Convert that DEM to vector format to select those steep slopes.

  4. Then it come to the way to determine the "Cut-off" line to get the slopes going downward to the features. I assume I need perform hydrology analysis on the DEM again. Should I do the Flow Direction next from the Hydrology Tools? And then Flow Accumulation?

Am I doing that correctly? Is there any step by step procedure can be provided?

  • What precisely have you tried so far in terms of tools and parameter values? That should help focus where you are stuck.
    – PolyGeo
    Commented Nov 14, 2017 at 0:45
  • I have already formulated the riparian area based on wetlands, forest cover, floodplain, Alluvial soils, wildlife corridor etc. What stuck me is to figure out the slope that go downhill toward Riparian Area, as well as the way to find a boundary that cut the slope that should be included and not. In doing this step it will need to perform analysis on DEM. But I am not sure what kind of methodologies should I use.
    – Kfg_0219
    Commented Nov 14, 2017 at 15:45
  • I wonder if the following link about creating watershed can provide some clues. gis.stackexchange.com/questions/116929/… I am not very familiar with hydrological analysis on GIS. If you can provide some step-by-step analysis will be greatly appreciated.
    – Kfg_0219
    Commented Nov 14, 2017 at 15:47
  • I have done the following so far: 1. Using the Spatial Analysis on the DEM to find out the slope. 2. Using relcass to find out the steep slopes that are more than 15%. 3. Should I do the Flow Direction next? And then Flow Accumulation?
    – Kfg_0219
    Commented Nov 14, 2017 at 17:15
  • Please use the edit button to provide any additional clarifications but take care not to invalidate any existing answers. You may or may not be better asking a new more focused question on a part that remains unanswered.
    – PolyGeo
    Commented Nov 14, 2017 at 19:50

2 Answers 2


You could use a slope-aspect interaction to determine slope direction in relation to slope intensity. A common metric in forestry application is the Stage (1976) sine or cosine transformation.

sca = θ cos⁡(α) or ssa= θ cos(α)
  where; θ = slope in percent and α = aspect in radians  

You then have a clear definition of the slope/aspect relationship that can be partitioned. Just keep in mind that zero slope has no transformation value. You can set these values to a very small number which will take the metric out of its expected range but clearly produce values associated with zero slopes. Here is an example of the metric(s) at a 50% slope across 10 aspects.

  Aspect    cosine     sine
  N         0.500      0.000
  N30E      0.433      0.250
  N45E      0.345      0.345
  N60E      0.250      0.433
  E         0.000      0.500
  ESE      -0.354      0.354
  S        -0.500      0.000
  SSW      -0.354     -0.354
  W         0.000     -0.500

I have an implementation of this metric, along with many others that may be of use, in the ArcGIS Geomorphometry & Gradient Metrics Toolbox and in the spatialEco R package, available on CRAN.


Stage, A. R. 1976. An Expression of the Effects of Aspect, Slope, and Habitat Type on Tree Growth. Forest Science 22(3):457-460.

  • Thanks Jeff. I will try to get the tool you developed as as part of the analysis.
    – Kfg_0219
    Commented Nov 13, 2017 at 20:28

Yes, you want to use the watershed tool in Spatial Analyst-Hydrology Tools-Watershed. After you finish determining the locations of your wetlands, lakes, riparian zones, etc. then combine those data, rasterize them, and reclassify them to either 1 = riparian areas to consider or NoData = areas not to consider. Then reclassify your slope data to 1 = steep enough to consider and NoData = not steep enough. Run the watershed tool using your riparian area to consider raster data as the in_pour_point_data in the watershed tool. The resulting layer will be ALL contributing watershed areas that flow into your riparian areas. Next use the slope reclassified layer to mask out only those areas that are steep enough to flow into your riparian areas.

  • Thanks GBG. Actually I already have the Riparian Zones (before Slope info) already, all I need to do is to work on the slope portion. Am I correct to figure out the Snap Pour Point after performing the Flow Direction & Flow Accumulation? Is that the right procedure? Thanks again
    – Kfg_0219
    Commented Nov 13, 2017 at 20:26
  • If you are using some extended river/lake/wetland/floodplain data set then snap to pour point would not be necessary since this tool just moves an inaccurate pour point to the lowest spot withing range.
    – GBG
    Commented Nov 13, 2017 at 22:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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