SDR project - catchment area

Hi, I have the problem that my watershed is not read in the SDR subproject, although that was not any problem with the seasonal yield subproject. Maybe you have an idea, what the problem could be? I have just uploaded the ordinary shapefile…

Maybe I should add, that in the user manual it is written that " * Watersheds (vector, polygon/multipolygon, required): Map of the boundaries of the watershed(s) over which to aggregate results. Each watershed should contribute to a point of interest where water quality will be analyzed.Field:

  • ws_id (integer, required): Unique identifier for the watershed."
    I had not put in the field “ws_id”, which would be a “1” in my case, and I do not know where to put it in, and I did not find the field in the sample data either. Cheers again. Ingrid
1 Like

Hi @IHartmann -

When you say that your “watershed is not read”, do you get an error? If so, please post a screenshot of the error so we know exactly what it’s complaining about.

If it’s complaining about missing the field ws_id, then you’ll need to add that field to your watershed shapefile in the GIS, and assign an integer value to it. If you look at the attribute table for the sample data file watershed_gura.shp, you’ll see the ws_id field.

I’m not exactly sure why SDR requires the ws_id field and SWY does not, but suspect that it’s a legacy from older versions of SDR that worked differently. I’ll see if that can be updated for future versions of InVEST.

~ Stacie

Dear Stacie,
ah, I have seen it now, it is in the attribute table. Will try that also with mine. Many thanks again. Cheers. Ingrid

OK, I have the results. But found, that they are highly dependent on the Borselli parameters, etc., which are only default values, so if these are changed, also results change substantially. Is that normal?

And I was also surprised, that I have a quite rich stream network in the SDR simulations, but only two rivers in the seasonal water yield results. Is that also ok?
Cheers again

Hello @IHartmann ,

Thanks for pointing out the ws_id field requirement! We will be removing this requirement in an upcoming release of InVEST since it is not used. That field was left over from a previous feature of the model that had been removed, and the ws_id field was inadvertently left alone.

I’ll have to let @swolny comment on the Borselli and other parameters and how the affect the outputs.

Are you running the models with the same threshold flow accumulation value and DEM? The TFA determines how many pixels must be upstream of a pixel to make it a stream. So, if the DEM and TFA value are the same between the two model runs, then the stream layers should also match.


1 Like

@IHartmann Yes, the Borselli parameters should have a significant effect on the results. They are typically adjusted when doing calibration with observed data, so the model has a better fit with reality. Do you have observed sediment data that you plan to work with?

~ Stacie

thank you! No, I do not have observed data…

Hi, I believe I have a similar problem with my modelling. InVEST does not accept my catchment area shapefile. The software also says that it can’t find the “ws_id” field in my vector, but I have already corrected it following the Gura sample data example (so, my vector does have the “ws_id” field).
So I am stuck. I can run the Gura sample data easily, but not my data.
What I imagine can be the problem is that my “catchment area” is actually a farm area that does not necessarily drains to a single point. Besides that, my rasters cover bigger areas than my catchment area vector. But I got no complaints from InVEST about these issues, only the missing “ws_id” feedback.
Can anyone tell me what is going wrong?
Thank you very much,

Hi @rearmelin, and welcome to the forum!

Can you post your watershed vector file so we can look at the fields?

The model does not judge whether the polygon is actually a catchment or not, but it does use it for aggregating your results. We highly recommend using an actual catchment, because otherwise you’ll be missing the influence of pixels outside of the farm area that still drain into the farm area and affect its water quality.

~ Stacie


Thank you for the help! Here it is. (2.1 KB)

Hi Stacie,

Thank you for your feedback. I thought about considering only the farm limits because we want to understand the potential impacts of the farm land use management in terms of erosion prevention.
Is there a better way to do that using the entire watershed? Can I compile the model results only within the farm limits? Maybe using SIG resources on the model`s final layers…
What would you do in a situation like this?
Thank you again,

PS: in case you could not get the file from the forum, here it is again. (2.12 KB)

Thanks for posting the farm shapefile. I do not see a field named “ws_id”, only one named “id”, so you’ll need to add a new field with the required name.

If you are only considering erosion prevention at the source, so, for example, you only really care about the USLE output of the model, then you can run the model and aggregate the results without a watershed, since USLE does not route sediment around the landscape, it just looks at what’s eroding from the pixel itself.

If you want to consider the role that farm management plays in reducing overall erosion to local streams, then it’s probably useful to include the larger catchment, since different management practices may help trap sediment originating from upslope of the farm. Similarly, different management practices are likely to affect downstream land use, which may need to trap additional sediment (if management practices lead to more erosion) or less (if management practices lead to less erosion). This way, you’ll be acknowledging that the farm is part of a larger catchment, with effects both upstream and downstream.

After running the model on the larger catchment, you can certainly zoom in to the farm to map and/or aggregate the differences in USLE/export/deposition that are caused by different practices.

~ Stacie

Hi Stacie, thank you very much.
I think I have a worse problem than… :slight_smile:
I just opened my shapefile in QGIS to check the fields. Please, see attached what I found. It is ok if you don`t want to go further with this, for this is probably not an InVEST issue.
But if have any idea about what is going on…
Thank you very much again!


I have no idea what’s going on. When I look at the shapefile you sent in QGIS, the attribute table looks like this:

What process did you use to add the ws_id field?

~ Stacie