Habitat quality threat layers

[hkarimi]

Hi everyone,
I am applying the InVEST Habitat Quality model to forested lands across the United States. One of the input layers represents forest disturbances, including wildfire, disease, and clear-cutting, etc., and is provided as a raster. My question is how the raster values should be encoded. Should disturbance types be represented by categorical labels, or should the raster contain numeric values?

[jesseG]

Hi @hkarimi ,

Thanks for your question about the Habitat Quality model. Please always refer to the User Guide for details on InVEST model data.

It states:

All values in [the land cover rasters] must have corresponding entries in the Sensitivity Table.

and

Each code must be a unique integer.

The rasters should have numeric values for each land cover class, but you’re also free to add additional fields with human-readable class/category descriptions.

-Jesse

[dave]

@hkarimi if you are intending to use that disturbance raster as a “threat raster”, then it should have values ranging from 0 - 1, representing the relative intensity of the threat. Perhaps all the pixels could be coded as 1 if all the categories represent an equal threat to habitat.

Data Needs - Habitat Quality — InVEST® documentation

[hkarimi]

Hi Jesse,

Thank you for your reply. It is not lad use layer, and as Dave mentioned it is threat layer that I think it should have value from 0-1.

[hkarimi]

Thank you, Dave. Yes, I understand now.

Hazhir

[hkarimi]

Hi Dave and Jesse, @jesseG @dave

First, thank you very much for your previous advice, it was very helpful. I have a couple of questions regarding accessibility and protected areas.

I am working at the U.S. scale and have downloaded the Protected Areas layer (PAD-US) for my study area. This layer includes four GAP status classes (1–4), where the level of protection decreases from class 1 to class 4. In the context of the InVEST Habitat Quality model, this suggests that these classes should be assigned different accessibility or threat-related weights.

I was wondering if you have recommendations for appropriate weights for GAP Status 1 and 2, which represent fully or strongly protected areas. For example, would values such as 0.1 (or similarly low values) be reasonable, or would you suggest a different approach?

Additionally, once the Habitat Quality model is run and produces values ranging from 0 to 1, what range would you recommend classifying as highly suitable habitat if I want to categorize the output map (e.g., low, moderate, high quality)?

Hazhir