Question about HRA results...

Dear all, 

Nice to meet you. I am excited to have started to use InVEST. 
It is quite an impressive piece of tool! 

I was willing to verify the calculation process and especially the spatially explicit risk score in shapefile, but I finally got a wrong result.

Here is the setting: 
- One habitat (seagrass)
- One stressor (fishing)

In preprocessor, here are the criteria that I have selected: 
- Exposure: Intensity Rating 
- Sensitivity: Change in Structure Rating
- Resilience: Recovery Time (★ this criteria was modified by spatially explicit risk score in shapefile)

As you can see in the attachment, the Recovery Potential (final ourput) has changed accordingly to the shapefile. It should mean that each patch of the habitat has a different resilience. 

But, when applying a stressor of constant intensity, the results in terms of risk are exactly the same. The habitat patch with low resilience should get a higher risk value I suppose. Or am I wrong?

I would appreciate if you could clarify this issue please.

Best regards,

Antoine 

1313 x 737 - 101K
1321 x 739 - 68K
1455 x 783 - 252K

Comments

  • jdouglassjdouglass Administrator, NatCap Staff
    Hi Antoine, could you send your inputs and logfile to jdouglass@stanford.edu so I can take a look?
  • Hi Douglas thank you for your answer sure I will send it right away!
  • kwyattkwyatt Member
    edited June 11

    Thanks for sharing your inputs. 

    I want to start by clarifying your criteria and inputs.  In your habitat ratings file “recovery time” (a component of resilience) is has a fixed value of 1 rather than varying spatially.  This differs than the images you included.  Intensity (a component of exposure) specifies “SHAPE,” which means that it varies spatially.  In the habitat ratings csv, I see that there
    is also a line for “trial,” which is being treated as a component of
    sensitivity. 
    We typically include multiple
    criteria for exposure, sensitivity, and resilience so it’s great that “trial”
    is in there, but since you didn’t mention it in your post, I wanted to make
    sure you knew it was one of your criteria.
      

    Recovery time is one of the final outputs and should, when
    used, be used alongside habitat risk. 
    Recovery time is the average score of only resilience criteria—those criteria
    listed in the CSV as “habitat only properties.” In your case, since you just
    have one resilience criteria and it does not vary spatially, the recovery time
    will match the score of 1 you gave “recovery time.”  In the inputs you shared, recovery time does
    not vary spatially and therefore neither will the output recovery time. 

    We more commonly focus on habitat risk, which not only account
    for resilience, but also sensitivity and exposure.  These results vary spatially based on the location
    of fishing activities and the exposure and consequence associated with fishing
    (including the spatially variable intensity). 
    If resilience varied spatially in the inputs then, yes, risk outputs
    would reflect this spatial variation.  In
    the results you shared, habitat risk (and ecosystem risk) vary spatially
    according to variation in intensity.  You
    may need to play with the symbolization to ensure variation is appropriately
    visualized.   You’re also correct that
    lower resilience (i.e. longer recovery time, etc) will have higher risk (to ensure
    you are ranking your criteria as you intended, please check the User’s Guide section
    for “exposure and consequence criteria in more detail”). 

    There used to a bug when spatially explicit criteria were
    used in conjunction with a buffer distance. 
    You may also try running your analysis with a buffer distance of 0 to
    make things are calculating results as you expect.  If it's all working right, you can add back the extra layer of complexity with a buffer distance. 

    I wonder if you may have shared a different set of inputs
    than the ones you describe.  Regardless,
    I hope some of this explanation helps to clarify your question.  Please be in touch if you have further
    questions.

     

    -       
    Katherine

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