Strange SDR Streams Classification Result

[dg34]

Hi all,

I’ve been using the SDR model in multiple watersheds and have come across a strange result associated with stream classification. For most of my watersheds, I’m using a pre-existing drainages raster layer to define the stream network paired with a very large threshold flow accumulation value. However, for some of my watersheds, no stream network is defined in the drainages layer, so I’ve been using a lower threshold flow accumulation value to manually define it.

I’ve experimented with a number of different values, and have found that for some, the model will calculate an empty streams layer, but report positive values for sediment export. Furthermore, when I pass the empty (all zeros) drainage layer, the model will generate a stream_and_drainage.tif layer that includes cells identified as streams/drainages, even though the output stream.tif and the input drainages.tif are both all zeros. When I remove the drainages layer as an input, the summary statistics are exactly the same, even though the output stream.tif layer is empty. Also, when I run the NDR model for the same DEM and watershed, the stream.tif layer is similarly empty, but no nutrient export is calculated. Any idea what might be going on here?

At lower threshold flow accumulation values, the stream.tif will include stream cells as expected, but I thought the strangeness of what I’ve been seeing was worth mentioning.

Thanks,
Dan

[swolny]

Hi @dg34 -

Well that does all sound strange. It sounds like you’ve done a lot of iterations, and I’m wondering if you could pass along inputs for a couple of the scenarios you’re describing, so I can look at the results more closely in a GIS. The easiest way to do this to make sure I’m running it the same way you are is to do this in Workbench (assuming you’re using the latest version 3.13.0):

1/ Within the SDR interface, click on Save As…
2/ Select Parameters and data
3/ Save the resulting .tgz file on your machine.
4/ Upload it somewhere and send it to me either here or at swolny@stanford.edu.

~ Stacie