SDR Model - Running Software Issue

Good day researchers, my name is Murilo and I am a master-student from University of Mato Grosso do Sul (UFMS). I’m developing a study about a Pantanal watershed, also I’m facing some difficulties about how to use SDR model properly (datasets, paramaters). I have some datas that are not working well in the software.

Therefore, I am interested on datasets (parameters) samples of SDR Model to improve my research.

Any thoughts and tips are welcome and I will be very grateful.

Have a nice week everyone

Best Regards.


Hi Murilo, welcome to the forum.

Have you explored the sample data for SDR that comes with InVEST? Find it in the sample_data folder of your InVEST 3.8.0 installation. And have you looked at the User’s Guide?

If you have specific issues with running the model, please start by reading Guidelines for posting software support questions


hi dave, how are u?

yes, i did, i ran the sample data into the model to see how invest works, but after files were generated into workspace folder, i couldn’t open any of this files, because they are not visible into any shape-program (i ve tested into ArcGis and Qgis) and both showed up a shape with watershed boundary content a black inside the boundary, i have no idea what happened because i followed the guide and the results are not what i expected.

Also, i have some doubts about biophysical table, how is the data input, when im interested in sediment ratios? (not the nutrient model as showed into data example)

It’s a little hard to say, but that sounds like it could be a problem only with how the GIS program is symbolizing the data in the results files. Adjusting the symbology rules (how colors map to values) could resolve that. Does watershed_results_sdr.shp have values in the attribute table?

If you are using the table from sample_data/SDR/biophysical_table_GURA.csv then it will have the required data for SDR, it also has additional data for the NDR model. The user’s guide has some more details on which columns are required by SDR:

yes, there are some values into attribute table, but how can i define colors map to a value range like any max-min value to each location instead a general value like showed into

Also, idk if it’s correct because idk how the results should be, but is this image right?


could i simply remove the load_p and load_n column (and add only the necessary data) from example biophysical table to use into a SDR model instead a NDR?

Hi @MuriloArima -

Yes, you can remove the columns in the Gura biophysical table that are not needed for SDR.

In both QGIS and ArcGIS you can change the symbology to correspond with different values in the shapefile. Each GIS is different, so I recommend looking at online tutorials to learn how. For example, here’s one that steps through vector symbology, including setting the colors based on a table attribute value. I’m sure there are many others.

~ Stacie

I have doubts about how much sediment is carried to the river, is there any possibility of invest estimate it? because my studied watershed is silting up and i wanna find out a way to estimate how much is going directly to the river.

Is there any info about river vector ?

thanks a lot guys

The sed_export results from the model provide an estimate of how much erosion arrives at the stream network, which is created from the input DEM. The model does not include any calculations within the stream network itself (like channel erosion or deposition.) The User Guide for SDR provides more detail about how this works.

~ Stacie

Hello everyone,

I’m currently working on the sediment delivery ratio model in some areas of Indonesia. However, during the running I always found an error.

Here I attached the model log file (.txt). If anyone could help me to understand what was wrong with the input model, it will help me very much.

Thank you!

ArmyandaInVEST-Sediment-Delivery-Ratio-Model-(SDR)-log-2020-05-15–08_56_03.txt (645.2 KB)

Hi @armyanda92, could you post your issue to a new thread so we can try to keep better tabs on things? Thanks!