SDR sediment model validation

My study areas are located within Piedmont region on central Maryland with all watersheds with around 1 square miles of drainage area. I don't have observed data available for model validation. Please advise reasonable course to adopt in such situations.


  • swolnyswolny Member, NatCap Staff
    Hi Ibraheem -

    If you do not have observed data, then we recommend using only relative values when reporting your results. For example, instead of saying that the model predicts a reduction of 2000 tons of sediment between two scenarios, you could say that it predicts a 10% reduction. You can also use the spatial patterns of the results to show which areas or sub-watersheds are providing more or less erosion control service, even if you are uncertain of the precise quantity of service provided by each sub-watershed. And it's always important to be clear with your audience that the model is uncalibrated/unvalidated, so they understand the level of uncertainty in the results.

    ~ Stacie
  • IbeIbe Member
    Thanks, Stacie.

Sign In or Register to comment.