Map extent of output different from input

I am using the carbon model and have an issue with the output maps (npv_fut, tot_c_cur etc). The tool ran fine (I had to adjust some errors along the way but these were fixed) but when I add the maps to ArcGIS Pro or QGIS the output maps do not align with the input map. They have the same projection, cell size, ncols, nrows, and the only thing is different is the extent.

Is there any reason why this would happen? And are there any suggested fixes?


HI @JBAronson ,

Do your current and future land cover maps overlap and have the same extents?



Hi Doug

Thanks for the reply. Yes they do

HI @JBAronson ,

That’s pretty strange that the extents would be different since those are the only two spatial inputs. What version of InVEST are you running?

If you can upload your data inputs I’d be happy to reproduce this on my end and investigate further. If they’re small enough you could zip them here. Or you could share via google drive or another cloud sharing service.



Hi Doug
I have added the files to google drive:
I appreciate you taking a look as I am not sure what the issue is.

All the best

Thanks Jon,

I just requested access for those files.


Could you share the other inputs as well so I can run it on my end?


carbon_pools.csv (354 Bytes)
Hi Doug, attached are the carbon pool data :slight_smile:

1 Like

Hi Jon,

I noticed that the raster output by the model changes the projection from the inputs.

Going in the projection is: EPSG:22832 - Douala / UTM zone 32N - Projected

Coming out the projection is: EPSG:3119 - Douala 1948 / AEF west - Projected

I don’t think I’ve ever seen this behavior from GDAL before. I’ll have to spend some more time trying to figure out what’s going on.


Hi Doug

Thanks a lot
I tried to reproject the output files to EPSG:22832 - Douala / UTM zone 32N but that did not work so the issue persists.

Hi @JBAronson ,

Following up that we’ve made an issue for the bug you’re experiencing. My guess is that it is an internal GDAL, PROJ library projection edge case. But you can check the link below for progress on that:



@JBAronson as a workaround you could try transforming all your inputs into a different projection to start with. For example, I would try EPSG:32632 - WGS84 / UTM Zone 32N.

I do see a notice that epsg 22832 was deprecated due to a datum error: Douala / UTM zone 32N: EPSG Projection -- Spatial Reference


Thanks Dave! Good catch and a good idea to try a UTM projection.