No shapely geometry can be created from null value (InVEST CV)

Hi, my name is Nonny from Indonesia.
Currently, I’m working on the vulnerability index along the Indonesia coastline with the InVEST model.

However, I found error warnings during the simulation. Here I attached the model log file (.txt).
I’m using the 3.8.9 version.

Thank you for the help!

Best,
Nonny

InVEST-Coastal-Vulnerability-log-2020-10-15–19_21_14.txt (22.9 KB)

Hi Nonny,

Thanks for including your logfile! Based on this, it looks like the model isn’t handling the landmass vector correctly, which is something that we should fix. Could you send me all of your inputs so that I can try to reproduce and fix this issue? If your inputs are small, feel free to attach them here to this thread, or else feel free to email me (jdouglass@stanford.edu) if that’s easier. If the inputs are large, it’ll be easiest to link to a google drive or dropbox folder, or something similar.

Thanks!
James

Hi James,

Thank you for your answer. I really appreciate it.
I try to run the model separately using the small study areas with the same landmass data, and it works.
So perhaps I will run the model separately instead of running it with the whole area.
Do you have any thoughts on it?
Or do you still need my input data? If that so, I will send it to you by email.

Many thanks!
Nonny

I’m glad you were able to run the model!

If you’re willing to share your inputs, that would be really helpful, so we can improve how the model handles the error that you originally experienced and prevent this issue from happening in the future.

James

Hi @jdouglass,

I seem to be having the same error with the landmass polygon, I was wondering if you have had a chance to look into this and know what might be causing it?

Thanks,

Meg
InVEST-Coastal-Vulnerability-log-2020-10-21–11_26_05.txt (8.0 KB)

@MegC, this error is most likely caused by a landmass feature not having any associated geometry, but could you send me your inputs for the Coastal Vulnerability model so I can take a closer look? If the inputs are too large to attach to this discussion, feel free to email them to me: jdouglass@stanford.edu.

Thanks!
James

Hi @jdouglass,

Thank you very much for taking a look into this, I have put the data into a google drive here: https://drive.google.com/drive/folders/1s_AqMJ2D1ZOb9Ut8h74Wnx1LtVVx3V7P?usp=sharing though the mangrove extent, AOI and human population data are the only ones which differ from the data given with the model. The landmass feature is the one which comes with the InVEST download. We also include geomorphology in our model, but this is currently unpublished so not included in this folder.

Thanks,

Meg

Thanks for your data! I’ve been running the model (minus geomorphology) on your inputs and am currently unable to reproduce this issue using the latest development build. Could you try running CV on the latest development build, linked below and let us know how it goes? I have a suspicion that this might have been fixed with an update in one of the libraries that InVEST uses for geometric operations.

https://storage.googleapis.com/releases.naturalcapitalproject.org/invest/3.8.8.post541+g0b694c4f/InVEST_3.8.8.post541+g0b694c4f_x64_Setup.exe

Thanks!
James

Hi @jdouglass,

Thanks very much for looking into that! I’m currently using version 3.8.9_x86 but will try download that and see if it works (will have to wait till tomorrow as I dont have admin rights). I will let you know how it goes! I was also wondering if you think the projection of the AOI is suitable for the model, as its equal-area rather than equidistant?

Thanks,

Megan

Hi @jdouglass,

Thanks for your help, that version of InVEST seems to run further and has been able to create shorepoints. However I ran into a new error, saying the WaveWatchIII points dont intersect the AOI. But I am not sure why this is the case? Have attached my log file, and any help would be appreciated!

Best,

Meg
InVEST-Coastal-Vulnerability-log-2020-10-22–10_12_09.txt (9.9 KB)

Hi @MegC, Thanks for pointing this out. This message is somewhat misleading because it really means, at least one of the shore points in the AOI does not have any WWIII data anywhere near it. The model searches for WWIII data within 3 degrees of each shore point. We should update that message and also consider skipping those shore points instead of raising an error.

As a workaround, you’ll need to exclude areas from your analysis that are very far away from any WWIII points. Ideally you would do that by cutting those areas out of the AOI. Another option is to edit the landmass polygon to remove those parts of landmass entirely.

All that said, it’s not obvious from looking at your input data where the problematic shore points would be. So I’m running the model to try to reproduce this and pinpoint the problem. I’ll post back when I learn more.

1 Like

Hi Dave,

Thank you very much for looking into this, it’s very appreciated. It would be really interesting to be able to run the model globally for the whole mangrove extent!

Thanks,

Megan

Update: Actually it was staring me in the face the whole time. The north coast of Africa and the Arabian Peninsula are too far away from the WWIII data to be included in the AOI. The purple are the WWIII and the yellow are the shore points that the model created along the landmass polygon.

You could cut a hole out of the AOI in order to exclude those regions from the analysis.

3 Likes