Nutrient retention service - NDR model

Hi everyone,

I am working with the NDR model to map the retention service... Reading a previous answer in the forum, I understand that the NDR factor could be used to map the retention value per pixel.

I runned the NDR model using both phosphorous and nitrogen retention and both rasters "ndr_n" and  "ndr_p" were exactly the same... I do not know if it is a good ideia to use identical rasters ("ndr_n" and  "ndr_p") to map both phosphorous and nitrogen retention. Is that the best option?

Would be better if I use the previous version of nutrient retention model to map nutrient retention service? Or is it more adequate if I use the NDR model (v.3.3.3)?



  • PerrinePerrine Moderator, NatCap Staff
    Hi Fernando, 

    The maps for N and P delivery will be the same if you use the same values for eff_p and eff_n, as well as the subsurface parameter.

    Retention at the subwatershed-scale can be calculated as the ratio between nutrient export and nutrient sources (the total "loads"). There's no pixel-scale retention output.

    We do not recommend using the previous versions of the model since the pixel-based output values are difficult to justify scientifically.

  • fmrfmr Member
    Hi Perrine,

    thanks for your answer. I have mapped the retention service as the ratio between nutrient export and total load (export/load) for each subwatershed. Indeed, that map is similar to the output raster ndr (ndr_n and ndr_p). 

    Nonetheless, using these approaches (export/load or ndr rasters) areas where higher proportion of nutrients are exported are associated with higher values, right? Thus, if I understood well, these areas do not seem to be the most important to retain nutrient. Am I missing something? I am planning to map the service of nutrient retention provided by native vegetation and indicate areas that are most important to conserve because of the provision of that service.

  • PerrinePerrine Moderator, NatCap Staff

    You are right, the high NDR values are areas that are exporting, not retaining, large amounts of nutrients.
    Like I said, there's currently no pixel-level outputs of the nutrient retention. One useful metric, at the pixel level, is n_export, which will show you areas of low or high export (and you'd want to conserve areas of low export). It's a different concept but useful nevertheless.

  • fmrfmr Member
    edited July 2017

    Hi Perrine,

    thanks for your suggestions. Yes, using “n_export”
    seems a good idea. I also tested two other approaches and I would like to share
    with you and see if they make sense.

    Considering that high NDR are areas that export
    large amount of nutrients (and that low NDR are areas that export low amount; "export
    = load * NDR"), I mapped “1 – NDR”. Doing that, high values are areas that retain
    high proportion of nutrients, and then I would want to conserve.

    I also estimated the proportion of the load
    that are retained in the landscape, using subwatershed data. I did that as
    "(load – export)/load". Both approaches generated maps similar and
    seem consistent. I am planning to use one of them to map retention service and I would be glad to hear from you.



    Post edited by fmr on
  • PerrinePerrine Moderator, NatCap Staff

    The 1-NDR value provides useful information since, as you note, it's the proportion of nutrient that will not be delivered to the stream. Remember that this is the retention from the N load on the pixel only: it doesn't account for the contribution of the pixel to retaining nutrient loads from upstream pixels.

    The other value, (load – export)/load, equals to: 1 - export/load, which is: 1 - NDR (because export/load = NDR).
    So you should obtain the same map indeed.



  • fmrfmr Member
    Hello Perrine,

    thanks  for your answer!

    Things are clearer now.

    Thank you!

Sign In or Register to comment.