Habitat Risk Assessment Output Error

Hello,

I have been using the Habitat Risk Assessment model as a habitat risk assessment in v3.7. The model runs successfully, however, the summary statistics excel file is empty. Also, the rasters generated by the model do not show any variation in the habitats and show as empty squares in GIS software. I wanted to know if I can have some assistance with this. Model inputs are attached. Thank you!

HabitatRiskTest_comp.zip (509.1 KB)

Thanks for posting. I’m having trouble reproducing your model run with these input data. Could you upload the logfile from one of your problematic runs so that I can see exactly what input parameters you are using?

Yes!

InVEST-Habitat-Risk-Assessment-log-2019-06-14–11_28_51.txt (27.5 KB)

Thanks! Upon closer inspection it looks like there is probably an issue with the Apple_test2.shp stressor layer. When I view it in GIS, it only has features that are extremely small (less than 0.1 centimeters across!). The coordinates of that layer appear to decimal degrees even though the projection is defined in UTM meters. I would probably go back to your original datasource for that layer and inspect it closely before doing any further data processing with it.

Hi! I ran the model again using a different stressor layer to ensure that the coordinates of the layer and the projection are both in meters. I am having the same issue as I was before with the output from the model. I will attach the new stressor layer along with the logfile. Thank you!

InVEST-Habitat-Risk-Assessment-log-2019-06-18–16_11_53.txt (28.5 KB)

It looks like only the logfile came through in the attachment, but not the stressor layer. The log still makes me suspicious of the quality of the stressor layer though. It appears to have taken 6 hours to rasterize “Cultivated_crop”, which doesn’t seem reasonable. I’m happy to take a look at it if you can’t find anything out of the ordinary with that layer. There may be file size limits on attachments here, so you could always share a link to google drive, dropbox, etc.

Got your data, thanks! So I think there’s a couple issues still at play.

First, the issue I noted above – coordinates of layers appearing to be decimal degrees but the projection definition stating a UTM projection – that is an issue for the AOI and habitat shapefiles. I can see that if I look at the layer properties in QGIS, which shows me this:

CRS: EPSG:26910 - NAD83 / UTM zone 10N - Projected
Extent: -96.6394850000195333,40.3754399998292683 : -90.1400609996912863,43.5011960001858142
Unit: meters

Those extent coordinates look like longitude/latitude decimal degrees while the CRS is suggesting they should be interpreted by software as meters. The solution is probably to redefine (not reproject) the projection definition to the correct system for those decimal degree coordinates. Or backtrack to the original datasources.

Second, that Cultivated_crop layer does not seem to have the projection problem, but it is very large and unwieldy (it has > 32 million vertices). If the original source of that layer was a raster, it is probably best to provide the raster directly to HRA instead of a vector (as of version 3.7.0, HRA can accept raster inputs).

In general, it’s always important and worthwhile to inspect all the input data in GIS before running an InVEST model. Good luck!

1 Like

Okay- got it! I’ll get start redefining those and let you know if it works! Thanks so much!