Problem: Calculating "d_dn"

Hi, 

Please, running InVEST SDR 3.1.1 x86, i found this problem in the picture below:

sdr                                   INFO     calculating d_dn
routing cython core      INFO     visit_stack on stream distance size: 626221 (reports every 5.0 secs)

I waited about 45 minutes and nothing, probably infinite looping.

Could someone help me?

Thanks again.
1366 x 726 - 531K

Comments

  • RichRich Administrator, NatCap Staff
    That's a bug in routing that should be fixed in the latest version of InVEST.  Can you upgrade to 3.2.0 and try it again?  http://data.naturalcapitalproject.org/invest-releases/InVEST_3_2_0_x86_Setup.exe
  • Hi Rich,

    It worked. 

    Just a question. The results "sed_export" and "usle_tot" are in tons/ ha / yr ?

    Is this idea correct? 
    Ex.: The idea is that in all sediment lost in a basin ("usle_tot"), only a part of it reaches the river ("sed_export")? 

    Thanks again
  • RichRich Administrator, NatCap Staff
    The units on sed_exort and usle_tot are tons / pixel / yr, not ha.  And yes, the usle_tot can be viewed as the potential loss due to surface erosion and sed_export is how much is exported because of the biophysical properties of the land cover in retaining some of that sediment.
  • Thanks Rich

    The output values in watershed_results_sdr (dbf file), t
    he units on sed_exort and usle_tot are tons / ha / yr or tons / pixel / yr.


  • RichRich Administrator, NatCap Staff
    The shapefile outputs are in terms of tons/watershed.   By the way, the units of the output files are described in the user's guide here under the "Interpreting Results" section: http://data.naturalcapitalproject.org/nightly-build/release_default/release_default/documentation/sdr.html#running-the-model

  • Hi,

    I'd like to confirm whether "tons" is referring to North American tons or metric tonnes. I can't find this in the documentation.


    Thanks!
  • PerrinePerrine Moderator, NatCap Staff
    Hi,

    "tons" refer to metric tons, i.e. the SI units.
    The erosivity (R) and erodibility (K) layers should also be in SI units. Note that their units partly cancel each other, so it's possible to have both in US customary units, and then the outputs will be in North American tons per acre. To convert to metric ton by hectare you need to multiply by 2.242.

    A summary of unit conversion factors can be found in the Appendix of the RUSLE guide, attached.
    (Ref: Renard, K., Foster, G., Weesies, G., McCool, D., Yoder, D., 1997. Predicting Soil Erosion by Water: A Guide to Conservation Planning with the revised soil loss equation. USDA)

    Best,

    Perrine
  • Hi Perrine, 

    Thanks, but I'm still struggling with this. The RIOS documentation says that K-factor values between 0.01 and 0.6 are normal. However, I suspect that this might be the case when using the metric form (t h)/(ha N). The K-factor inputs for RIOS are asked to be in (t h)/(MJ mm). [On a side note, there is an error in the RIOS documentation that replaces "h" with "ha".]

    Following the Wischmeier nomograph method using the steps shown in "Use and misuse of the K factor equation in soil erosion modeling: An alternative equation for determining USLE nomograph soil erodibility values" by Auerswald et al (2014), I have produced a K-factor map for the country of Ethiopia with values ranging from 0.04 to 0.57.

    The units are (t h)/(ha N). One must divide by 10 to convert to (t h)/(MJ mm). For this reason, I either believe that the RIOS documentation recommendation of values between 0.01 and 0.6 is wrong. Or else, the units needed by RIOS are actually (t h)/(ha N).

    I hope one of these can be confirmed?
  • PerrinePerrine Moderator, NatCap Staff
    Hi jelliso, 

    I agree with your points, and apologies for the typo in the RIOS documentation.

    Your values of K in (t h)/(ha N) look good to me. As you point out, if you divide them by 10, you obtain value SI units in the expected [0.001-0.06] range.

    If you want to obtain values in US customary units, you need to divide them by 0.1317 (as per conversion .pdf inthe previous post).

    Now on to RIOS. I think the documentation hasn't been updated and the range of values should now say "0.001 to 0.06"). Note that it does not influence your outputs since all values are normalized in the algorithm (so the unit doesn't matter). 

    For InVEST, however, one should make sure to have SI units ( (t h)/(MJ mm)) to match the SI units of erosivity.

    Hope this clarifies the issue!

  • Thanks! All clear!
Sign In or Register to comment.