This forum is shutting down! Please post new discussions at

Habitat Quality Module

Hi NatCapers

1) I have a baseline and two future scenarios I want to consider. Should I run it three times with only one LULC map at a time, putting the TIF file in the "Current Land Cover" spot? Or once using the Current, Future and Baseline? (I have previously created different LULC maps for my two scenarios using the Scenario Generator.) What is best practice? Ultimately I want to contrast the outputs between the baseline and two scenarios.

2) I'm confused about linking the threats and sensitivities tables. In the threats, I want to stipulate that an existing LULC type is a threat to biodiversity because it is low quality habitat (this is based on literature and my own fieldwork), and in my future scenarios, it is an expanding LULC type. In the sensitivity table, I have to list each LULC type and list the threats. So, since the LULC type is both part of the landscape AND a threat, should I put 0 in the threats column since its not senitive to itself? 

Does that make sense? Am I conceptualizing of this module correctly?

3) For threats distance, does the model us/understand less than one - so can I use 0.25, meaning 250 km? Or does it round it down to zero? 

Thanks! I did get the model to run using the baseline map I have, so thats something!



  • RichRich Administrator, NatCap Staff
    Hi, I'm not an expert on this model, but...

    #1: The internals of the model run the HQ analysis for current, future, and baseline.  If you want to keep track in your head that one baseline is one future and the future is the second future, there's no harm in that.  Also no problem if you want to run it 3 times.

    #2: I'm going to see if I can get Nirmal to comment on this.  He's NatCap's active user of this model.

    #3: Yes, you can use floating points for distance.

    I'm going to ping Nirmal via email.  So more for you on #2.
  • OK, thanks. I understand how the model interps roads/points, etc but not how it uses LULC type data, so I would like to hear from an experienced user.

    I appreciate the help! 

  • RichRich Administrator, NatCap Staff
    I just got an email that he's out of the office until Aug 28.  Sorry, he's quite good at responding to these when he's at his computer.  So more soon!
  • swolnyswolny Member, NatCap Staff
    Regarding the 'baseline' input to the model, it is mainly there for calculating habitat rarity. If you include a 'baseline' map, you'll get a habitat rarity output that shows the difference in habitat occurrence between baseline and current LULCs.

    A few ideas for #2. In the sensitivity table there's a HABITAT column. You could use that to rate this particular LULC class as 0 or a very low value for it being suitable for habitat, based on your fieldwork. If, however, you want to say that not only is this LULC type unsuitable, but also specify a threat distance around it (as you might for something like, say, urban areas, where people could travel X distance to hunt or log) you could extract the particular LULC type that is a threat, and use it as a Threat raster input to the model and populate the THREAT table accordingly.

    ~ Stacie

  • It's been a while since I ran the model, but what Stacie said about point #2 makes sense. If one of the LULC classes is a threat to neighboring LULC classes, then it should be extracted into its own threat raster and you would need to specify its radius of influence. I *think* it should be ok if that threat also exists in as a LULC class with low or zero sensitivity to itself.

    However, since the threat itself is expanding in the future scenario, you would need to run the model a second time with the updated threat raster for each scenario, rather than running it once with current, baseline and future.
  • Hi all,

    So I took your advice and structured my threats and sensitivities as you suggested, extracting the threat LULC from the map and making it its own layer, and assigning it a sensitivity of 0. It seemed to work and produced the maps that "makes sense" to me. 

    Also, the online documentation for the model, there is instructions to make a landscape/biodiversity scores -- summing up the individual grids. Since the output is a float raster, I used the map calculator to multiple by 1000, then used the Int feature to convert it to whole numbers and add an attribute table. Then I summed up the totals for each scenario map and made then expressed them as relative changes in percents? Is that what you recommend? (Or did I do something bad....?)

  • While I have calculated the percent difference in scores in previous analyses, I'm not sure how informative that is (since the score is not pegged to any physical units, and so it doesn't mean that the habitat is x% better is one scenario than another). Still, I suppose it can help compare among different scenarios to rank order them for their overall habitat quality provision.

    It could be informative to do a raster subtraction between the different scenario outputs to look at the spatial variation in differences in habitat quality pixel scores across the landscape. It would help assess which areas had relatively gained or lost more quality across scenarios.

    But I'm not absolutely sure whether habitat quality pixel scores are directly comparable across scenarios (since the threat raster would have changed from one scenario to another). If only the LULC had changed, they would be comparable. I think it still should be okay, but it would be good to check the formula to make sure. Do others know?
  • Hi Nirmal, 

    We'll how this all works out. I wish there was a NatCap homework desk to check my work - no one else I know how used these tools. I did the HQ, Pollination and Carbon modules. The maps are great but giving a sense of relative change helps too I think.

    Thanks for the advice and help.

Sign In or Register to comment.