Some doubts in the nutrient retention results

Dear all,

I use the nutrient retention model in my study case.
I have a inland marsh of 200 ha that is a part of a little natural reserve of 817 ha.
I run the nutrient retention model for the watershed (12900 ha) of the river which passes through the inland marsh.
I am quite supraised about the results, in particulary about the nitrogen retention (and also the phosphorus) of the inland marsh that is near to zero. Is strange i think, also because it has the highest vegetation filtering value, 0.80.
I can see this in the n_retention and p_retention raster, and I think is in Kg/ha.
My inland marsh is situeted at the end of the river (and the watershed), just before the intersection with the main river.

someone can help me?
best regards
thank you.


  • PerrinePerrine Moderator, NatCap Staff

    Thanks for sharing these results. 
    This does not look like something we've observed before. On the contrary, pixels next to the stream (your inland marsh) usually show high retention values. To interpret your results correctly, remember that nutrient retention values are based on the efficiency value of the pixel and the total load coming from upslope. So if the load from upslope is small, the retention is small too. Details on the routing algorithm can be found in the User's guide.

    Also, something I would check (if you haven't done so already) is the stream layer. Does it match the LULC (ie. are streams correctly located in the marsh)? InVEST computes the flow accumulation and stream map based on the DEM and sometimes they don't overlap with the LULC.

    Hope this helps,

  • Hi

    first of all thank you for aswer me.

    I was looking my data and the river course deirived from the DEM is differet from "real" course of the river because during the years humans change the course of the river. So maybe could be what you say that the LULC does not match with the stream layer.
    But there is another fact: around the inland marsh there are permantent irrigated lands with channes derived from the DEM that does not exist in reality that retain more than inland marsh with filtring value of 0.25.

    thank you for the attention
  • PerrinePerrine Moderator, NatCap Staff
    Hi Lorenzo, 

    It looks like you have a human-modified landscape, which may impact your modeling (as mentioned, InVEST will assume that flows follow the topography whereas flow direction may be different in reality)

    I'm not sure I understood your second point. Let me know if you need more help!
Sign In or Register to comment.