This forum is shutting down! Please post new discussions at

Spatially Explicit Criteria


We are trying to use the HRA model to look at cumulative impact of human activities on flatback turtle nesting grounds. We have a question regarding spatially explicit criteria.


We would like to consider spatial intensity of the stressors. We describe below two different examples:


- Infrastructure (stressor is “Habitat loss”). This is a polygon layer. Can we assign a rating to each infrastructure, i.e. each polygon, based on its area? So for instance a port with an area of 2kmwould have a rating of 3 and a small building of 10m2 would have a rating of 1. We would like then to input this polygon layer into the model. Can we directly copy this layer into the “Input \ Spatially-Explicit-Criteria \ Exposure” folder?


- Artificial lights: this is a raster layer showing the intensity of light in each 400mx400m cell.

(1) Can we directly input this raster layer into the model or does it need to be turned into a shapefile first? i.e. can we directly copy this layer into the “Input \ Spatially-Explicit-Criteria \ Exposure” folder?

(2) If it needs to be turned into a shapefile first, then we understand that the model will convert it back into a raster layer. How do we then make sure that the model retains the resolution, attribution and extent of the original raster layer?

Thank you in advance for your help


  • kwyattkwyatt Member
    Hi Sabrina, thanks for your questions.  

    In regards to the infrastructure stressor, you can assign each polygon a unique rating for intensity (or any other criteria you feel would be appropriate).  Yes, you can include this in the exposure folder following the naming convention outlined in the user's guide.  The sample data offer a good guide for how to do this too. 

    However, I might suggest you think about whether using area to rate the intensity is the most appropriate.  The model already accounts for spatial overlap so you might create some accidental redundancy.   In the past, we've used spatially explicit criteria to differentiate different intensities or impact--e.g. greater invasive species density, greater fishing effort, or different kinds of boat traffic.  If there are different densities or types of infrastructure, those might be other ways to reflect variation in that stressor.  

    As for light intensity, you'll need to convert it to a polygon first.  If you want to maintain some of the variation in light intensity as a spatially explicit criteria, you'll want to think about how to divide the continuous light intensity into appropriate bins--quantiles have been a nice starting place for us in similar applications. 

    The model will convert the polygon back into a raster to run the model.  The extent will remain unchanged.  You can ensure the same resolution by setting the resolution in the HRA user-interface to match that of the original raster.  That said, depending on the resolution of your other data inputs, you may want to consider revising the resolution for the model.  

    I hope this helps and please reach out if you have additional questions. 
  • SabrinaSabrina Member
    edited May 2018
    Hi Kwyatt,

    Thanks very much for your helpful answer.

    Good point about the infrastructure stressor. We will have a think about a different unit we can use to measure the intensity of this stressor. 

    Light intensity: sounds good. Do we need however to keep the model rating scale of 1 to 3 even for spatial intensity ranking or are we free to use a different scale (eg quantiles)?

    Regarding the raster layer build by the model (i.e. raster_grid_base.tif ), how does the model proceed to build it, i.e. what is the starting point and in which direction does it proceed? I would assume this is important to know as that might influence how the area of interest is divided into cells and therefore might influence how polygon layers are converted into raster layers by the model.

    Last question if you don't mind: Is there a way to take into account the spatial intensity
    of the habitat itself? In our case, we have more than a hundred nesting
    beaches, each with a different turtle nesting density and we would like to be able to
    take that into account in the model. Indeed, a low density versus a high
    density nesting beach might not have the same resilience to certain stressors. Can we assign a rating to each polygon/beach based on turtle density and include this layer in the folder 



    Post edited by Sabrina on
Sign In or Register to comment.