My model ran really fine, but when i open the base flow output there are random places with no data. I ran the model with almost the same input data but differente LULCC scenario and worked fine.
But now somehow i am having NoData in some parts of the watersheed and also my ouput shapefile Aggregated results has no values for qb and vri_sum. What could be the problem?
Hello @clara, and welcome to the forum!
Do the NoData areas appear to be related to any of your model inputs? Anywhere any of the model inputs have NoData cells the output results will also have them. If the only thing that changed is the LULCC map, what is different between the scenario and baseline maps?
If it doesn’t appear to be related to NoData in your inputs, it would be helpful if you’d post an image of the output, so we can clearly see the location of missing data, and perhaps a comparison with the input LULCC maps that changed.
~ Stacie
Hello @swolny thank you for your reply!
Not really, i double check several times to see if the NoData areas are related to some of the input data, or some classe from LULC but not really. It does not overlay exactly with any pattern from the input data. I was working with the InVEST Workbench of 3.14.1 and because of this error i downloaded the new workbench version 3.14.2 but even like this is not working properly.
I will post here the image of the Baseflow output raster to show the Nodata. As i said , not only the baseflow is with error but also the Aggregate results shapefile.
I changed the scenarios, in this case, other LULC but i have the same problem, the NoData in the same places.
It might be easiest if you can share your input data so I can try it out. Here are instructions for how to package up your model run in a way that makes it easy for me to duplicate your parameters. Then you can share it through any cloud platform that works for you. Thanks.
~ Stacie
Thank you!
here is the drive link with the file saved from InVEST with the Parameters and data:
Thank you in advance for the help.
Hi @clara -
Thanks for the data, I ran it, see the same areas of missing data, and don’t understand what’s causing it. It does remind me of an old bug with our sediment model, where using a different DEM fixed the problem. So if it’s easy for you to try a different DEM, it would be interesting to see if it fixes this baseflow issue. Meanwhile, I’ve asked the software team to look into it more deeply, and we’ll let you know what we find.
~ Stacie
Hi @swolny ,
I tried with other DEM and i still have the same problem. Hope you guys can help me! It is for my master thesis.
Hi, @clara
I’m experiencing the same issue: the aggregate results shapefile is missing values, and the VRI shapefile isn’t appearing. I’ve tried switching the DEM raster, using both older and newer versions of InVEST… but nothing has worked so far. Please let me know if you come across a solution 
@gabbmdn The same for me. I tried with different DEM. The first time worked fine… now for other scenarios, using LULC of different years is not working anymore. I hope we get a solution soon.
some news about the bug?
We’re working on it but don’t yet have an explanation for you.
Hi @clara , after careful troubleshooting we have concluded that the source of the problem is missing data in the precipitation rasters. There are two problems:
L_sum_avail.tif). These missing precipitation values propagated through the local recharge calculations to all the other layers where you observed missing data.Please make sure that all your input data have a defined nodata value. For best results, do not use nan as the nodata value.
Also make sure your precipitation layers cover all the valid DEM pixels. It is not necessary to clip/mask the precipitation rasters to the watershed before running the model. The model will handle this appropriately.
Thank you,
Hi @dave ,
I encountered the same issue as @clara and tried the suggested solutions. They worked for the baseflow rasters, but I’m still unable to generate the VRI raster. Do you have any insights into what might be causing this problem?
Hi @gabbmdn , thanks for trying out that solution. My guess is that it’s still a problem of undefined nodata or nan in one of your input layers, but I can’t be sure. If you would like, you may share the latest version of your input data and I can take a look.
Thank you for your help, @dave and @swolny
After considering your answer and feedback, i set NoData values for Pr and ET0 rasters but i am facing the same problems with the random NoData values on Baseflow raster and no results for the aggregated shapefile.
Here’s an update on my progress and the challenges I’m facing:
I successfully modeled the LULC 2023 scenario 2 months ago, but now after almost two months of effort, I have been unable to get reasonable results for other scenarios, such as LULC 2000. The discrepancies make it difficult to draw meaningful comparisons between the scenarios, which is critical for my analysis.
Steps I’ve Taken:
NoData Settings:
Current Issues:
I am sending in the next message the parameters and data of this Test for 2000 and for the test of 2023 that worked out.
Hope you guys can help me.
Hi @clara , thanks for sharing your data again. The zip file only included a json file, not the actual data. If you’re using the Workbench, please make sure to select “Parameters and data” from the “Save as” menu. Thank you,
@dave here is again the link of my drive folder with the data from the 2000 SYM model ( parameters and data)
https://drive.google.com/drive/folders/1Wroj4pKzeVsfamXgD37wrpi81nMLhZ_7?usp=sharing