NDR model result with empty pixels

Science, Research & Data
1

My NDR (Nutrient Delivery Ratio) model output shows no pixels within my study watershed (all masked out).

  • I used the filled Copernicus DEM (GLO-30) and delineated the watershed using the InVEST Delineation Tool.
  • For land cover, I used Global 30 m Land-Cover (GLC_FCS30D), which I reclassified to create the LULC input.
  • Although the DEM and LULC have the same coordinate system, their pixels did not align exactly, so I used the SNAP tool (based on the DEM) to match the land cover pixels with the DEM.
  • I used total precipitation as the runoff proxy.
  • The biophysical table was carefully prepared based on the literature.
  • Subsurface Critical Length (Nitrogen) and Subsurface Maximum Retention Efficiency were both set to zero as subsurface flow of nitrogen is not considered in my study.

Despite these steps, the NDR result shows many pixels. The colorful areas you see from the the next picture are from an underlying land cover layer . only black-and-white color pixels represent NDR model result .

I am not sure where the issue lies. Could you help me troubleshoot why the NDR output is empty?

image
image1640×772 250 KB

2

Hi @Eshetu -

Which version of InVEST are you using? When you have a problem, please always upload the log file (.txt file with a long name) that is created in your Workspace, which provides us with useful information like the InVEST version.

This reminds me of this post where someone had a lot of missing data in their results. NDR uses some of the same processes as SWY, so could have similar problems.

Please check out that post, see if it helps, and post your logfile if it doesn’t solve your problem - thanks!

~ Stacie

3

@swolny

Thank you so much for your help. The issue I encountered with InVEST 3.16.0 was resolved after updating to version 3.16.1. Now, only the streams show “no data” values. Is it correct to proceed with the analysis in this case?

I have another question. All numeric entries in the biophysical table were derived from recent publications, but those studies did not include a “load_type_p” field in their tables. As there is no load-type data available in previous studies, I used “measured runoff” under the column “load_type_p,” following the format provided in the user guide sample table. What is your advice regarding this approach?

Thank you once again for your guidance.

log file is as follow:
InVEST-ndr-log-2025-07-24–12_27_26.txt (48.5 KB)

4

Glad to know that the issue was resolved, thanks for letting us know.

Yes, the freshwater models generally assign NoData to streams, since no in-stream calculations are done, so you can proceed with the analysis.

Your other question about nutrient loading type is a duplicate of this post, and I will answer there.

~ Stacie

5

@swolny : Thank you so much for your prompt response and for your recommendations regarding my problem.