Habitat Risk Assessment (HRA) model issue

Hello Everyone,

I have a question about using files containing multiple polylines as stressor layers. I am using the Habitat Risk Assessment model to evaluate the potential impacts of shipping on whale populations. My stressor data are aggregations of ship track polylines that are divided into three separate shp files by ship type (i.e., bulk carriers, cargo, tanker).

The problem arises when the model calculates cumulative risk in the whale habitat. I believe it is only accounting for the overlap in stressor layers rather than the intensity of polylines within a layer. For example, if my tanker shp file demonstrates that 100 tankers passed through the exact same area, the HRA model gives that the same weight as any area where only one tanker passed. The additional risk would only be accounted for if, say, a bulk carrier and a tanker passed through the same area (i.e., two overlapping stressors).

The only solution that I can think of is to create separate shp files for each ship track polyline. However, given that I have hundreds of ships, this is infeasible. Does anyone have any suggestions for ways to incorporate the spatial variation in the intensity of shipping within the HRA model? 


  • Thanks for your question.

    First, can you clarify that your shipping lanes are represented as polygons rather than polylines?  The model expects polygons.  

    The model assumes equal intensity for each stressor unless otherwise specified.  To specify varying levels of intensity you need to use spatially explicit criteria.  There's additional information about this part of the tool in the User's Guide and there's an example in the sample data.  Briefly, you create an additional folder for spatially explicit criteria in which you can insert shapefiles that specify, using additional attribute information, spatial variation in intensity (or any other criteria) of a given stressor on a given habitat.  In your example, within a single shapefile, some of the shipping lanes could be assigned an intensity of 3 and other lanes an intensity of 2 or 1.  Keep in mind that the model uses categories for each criteria so you won't be able to specify continuous variation in intensity.

    I hope the spatially explicit criteria feature in the tool helps address your issue!  As you explore this feature, please feel free to reach out if you have further questions. 

    - Katherine 
  • Hi Katherine, 

    Thank you for your comments. My ship tracks are in polyline format at the moment. Could you clarify where in the HRA model documentation you saw that it expects only polygons as stressor layers? I am having trouble finding it. Also, the model is currently functioning using polyline shapefiles as stressor inputs -- do you know how it is currently handling the polylines? For example, are the lines internally converted into polygons? 

    Do you have any suggestions for the best way to transform the polylines into polygons? Perhaps use polygons created from the buffered lines? Or rasterize them into polygon cells with intensity values based on different shipping intensities in each cell? 

    Thanks again for your help -- I and my colleagues really appreciate it! 

  • Hi Kelly,
    I am doing something very similar to you and also had no idea that the stressors had to be polygons! What a headache..
    What I've done with my vessel tracklines is create a raster of the lines using one of the density tools, then depending on what information you want to convey using Con or Reclassify to change the unique values of the cells into category groupings (like, 0, 1, 2, 3), and then run Raster to Polygon. For HRA you need to add a field into that shapefile called 'Rating' and then populate it with the weights (0,1-3) you want to assign based on your criteria. For example, if you can generalize your ship patterns into low, medium, and high density then you could use the spatially explicit criteria to make the low density area a score of 1 and a high density area a sore of 3 on the intensity criteria. That way invest wont treat all areas with vessel presence as having the same response. Also make sure to follow all the other rules about using spatially explicit criteria like the folders and naming conventions
  • Hi rudebuschjane, 

    Thanks a bunch for the work around! I'm glad I wasn't the only one facing this problem. I had thought about doing something similar and have actually rasterized my tracklines for another project, but I worry about dividing the ratings into just 3 density categories -- especially given that it's a heavily right-skewed distribution of intensities with many areas experiencing little to no traffic and a few areas experiencing a lot. However, it seems like using the spatially-explicit criteria may be the best way to get the HRA up and running. 

    Best of luck with your analyses! 
Sign In or Register to comment.