SDR model: help to understand very low output values

RaquelSobrinoRaquelSobrino Member
edited July 19 in Freshwater Models
Hi there again,

I'm running the SDR model to get a portfolio of ecosystem services. The model runs apparently perfect, but I get very very low values. I've tried different analysis, among them:

For Borselli K parameter = 2
  • TFA=1000  gives me sed_retentsdr.shx=0 and sed_exportsdr.shx=0 in each subwatershed.
  • TFA=16 (which produces a flow_acumulation.tif quite similar to what should be) gives me sed_retentsdr.shx=3x10^-6 to 8.9x10^-5 and sed_exportsdr.shx=1x10^-6 to 8x10^-5 depending on the subwatershed.
Changin Borselli K parameter to 1:
  • TFA=1000  gives me sed_retentsdr.shx=0 and sed_exportsdr.shx=0 in each subwatershed.
  • TFA=16
    (which produces a flow_acumulation.tif quite similar to what should be)
    gives me sed_retentsdr.shx=3x10^-6 to 7.7x10^-5 and
    sed_exportsdr.shx=1x10^-6 to 6x10^-5 depending on the subwatershed.

I've also tried several erosivity values, because there is no data from the area, and using different formulae R changes enourmously. It seems to be the parameter that most effect has in the output values, but there are still low.

Any suggestion what can I do wrong?

Thanks for your help

Raquel

Post edited by RaquelSobrino on

Comments

  • swolnyswolny Member, NatCap Staff
    Hi Raquel -

    I would first check all of your spatial inputs to make sure that they have reasonable values and are in the correct units. It's interesting that a TFA of 16 creates a good stream network and a TFA of 1000 gives 0 results - are you using a DEM with large cell size? 

    ~ Stacie




  • HI Stacie,

    thanks for answering so fast.

    My DEM is 30 m, is it good enough?

    For the erodibility (k) raster I follow the instructions in the document you posted https://forums.naturalcapitalproject.org/uploads/FileUpload/1b/73867fc6231e8ff8507871f5bd33ae.txt.

    And for the erosivity index I used different equation from Renard and Freimund (1994). Because they were derived for different places (non of them similar to Rimac watershed) I get a set of values ranging from 6000 to 1300000 aprox. This is the parameter that most effect have, but still my outputs are very low.
    Raquel
  • Hi Stacie,

    Just to let you know that I finally managed to solve the problem. One of the raster hadn't been correctly projected and the cell size was incoherent with respect the other raster.
    Cheers
    Raquel
Sign In or Register to comment.