I am wondering if there are better methods for determining the mean watershed slope, other than taking the mean of the slope values contained within the watershed?

For example, could or should one take the main channel slope as the watershed slope? Or the average of all identified channel slopes as the watershed slope?

I am looking to categorize several watersheds using, among other things (i.e. basin compactness, etc.), mean watershed slope.

Edit: I am assuming an estimate of watershed slope can act as a predictor for streamflow characteristics. I do not want to actually estimate the streamflow, but group similar basins based on several attributes that might be used to predict streamflow.

  • What are you trying to achieve? Are you interested only in how steep the watercourse is or the surrounding area? What impact does the surrounding area have on the watercourse and how does that affect your study? For example, in a sediment study steeper areas are less likely to be high sediment areas irrespective of how flat the water is, but a soil study would be a better indication than slope. Flow rate would mostly be impacted by the grade of the watercourse and not so much the surrounding terrain.. Jul 1, 2015 at 3:25

1 Answer 1


It looks like statistical model. If yes, mean slope derived from slope raster is the best because:

  • after spending a lot of time computing others, you'll find a very good correlation between them
  • raster based is easiest of all.

On the other hand if ir is about floods, i.e. max values, you might want to combine some independent variables into one. E.g. it is common practice to derive time of concentration by using flow length and slope, and this time around it is a longest flowpath and slope along it.

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.