NDR Output: Conceptual Understanding

[atershy]

Hello InVEST users!

I have a couple of queries about how to interpret the outputs of the NDR model. I hope that someone can help shed light on either of the following.

1. I’m using the NDR model to understand how effective riparian vegetative buffers are for nutrient retention in my local watershed. How can I determine exactly how much N ( in kg) the riparian buffers retain? I have a shapefile of all the riparian buffers in the watershed. I was hoping to use the shapefile to extract values from one of the output rasters. Is there an output raster or a combination of outputs that will tell me the amount of N retained by each pixel?

2. What’s the practical difference between “load_n” and “modified_load_n”? I understand that the “modified_load_n” takes into account the amount of runoff transporting nutrients. But, I don’t understand why, for a given pixel, my “modified_load_n” value is greater than my “load_n” value. Is it because the modified load of a pixel is the accumulation of load from other pixels carried to that pixel by runoff?

[jesseG]

Hi @atershy,

Thanks for your great questions about InVEST’s NDR model outputs.

1. Please do bear in mind that unless you have compared, calibrated, and validated the quantitative results against empirical data from your study area, we strongly discourage placing high confidence in the absolute values of the NDR model outputs and instead only recommend interpreting results as relative across space (comparing pixels values to one another). This will give you an understanding of where export, retention, etc. is higher and lower across the landscape. This is explained in the InVEST User Guide, here.
   That disclaimer aside, to get at the potential supply of nitrogen retention services provided by vegetation, use the “modified_load_n” raster result MINUS “n_export.tif”. That will give you annual nitrogen retention on each pixel in kg.

2. ‘Modified load’ will be higher than ‘load’ for any pixel that has a higher nutrient runoff proxy (potential) than the average runoff proxy of all the pixels in the raster. This is explained in the User Guide’s section on Nutrient Loads.

I hope you find his information useful. Please let us know if you have more questions or would like further clarification.

-Jesse

[atershy]

@jesseG

esults as relative across space (comparing pixels values to one another). This will give you an understanding of where export, retention, etc. is higher and lower across the landscape. This is explained in the InVEST User Guide, here.
That disclaimer aside, to get at the potential supply of nitrogen retention services provided by

Hi @jesseG ,

Thank you very much for the explanations! I am in the process of calibrating and validating the model results with local empirical data. I will make sure to do so before interpreting any absolute values of N retention.

Cheers!