Hi everyone, I’d like to enquire about how to calculate habitat degradation caused by each threat.
In my habitat-quality model, I have set 5 kinds of threats and get a map of total habitat degradation caused by all the 5 threat sources.
Now I want to calculate how much degradation is caused by each threat. I would appreciate if you could give some advice on it.
What do you expect to happen?
According to the user guide, total habitat degradation is calculated as the formula shown in the following screenshot.
Dxj = ∑(r=1;R) ∑(y=1;Yr) [ (wr/∑(r=1;R)wr) * ry * irxy * βx * Sjr ]
In my point of view, the total habitat degradation is actually the sum of degradations caused by each threat. So I guess the habitat degradation cause by each threat is calculated as : ∑(y=1;Yr) [ (wr/∑(r=1;R)wr) * ry * irxy * βx * Sjr ]. Is this correct?
What have you tried so far?
In the intermediate folder, I got :
(1) filtered_ThreatMap, for each threat. User guide says they are threats layers processed with Threats data table attributes.
(2) sens__ThreatMap, for each threat. User guide says they are sensitivity applied to different threats.
I’m not sure how filtered_ThreatMap is calculated. Does each filtered_ThreatMap is calculated as : filtered_ThreatMap = ry * irxy * βx ? And does sens__ThreatMap = Sjr?
If the answer is positive, is it correct to calculate the map of the habitat degradation cause by the i_th threat as:
filtered_ThreatMap_i * sens__ThreatMap_i * (weight_i / the sum of all threats’ weights)
(weight_i is from the Threats data table attributes)
Thank you for your time in advance, and looking forward to hearing from you soon.
Yes, the degradation for a single threat r can be calculated as
Which is equivalent to:
β , the accessibility factor, is access_layer.tif
is the relative weight of the threat
Sⱼ, the sensitivity factor for habitat j, is sens_<j>.tif
r is the input threat raster
, the distance-weighted threat raster, is filtered_<r>.tif
The total degradation for all threats Dₓⱼ is output as deg_sum.tif. The model does not create rasters for degradation from individual threats Dₓⱼᵣ, but you could calculate them by multiplying: access_layer.tif x relative weight x sens_<j>.tif x filtered_<r>.tif.
I would appreciate it if you could check whether my understanding is right or not:
access_layer.tif, sens_.tif and filtered_.tif are all in the intermediate folder.
Relative weight could be calculated by myself from weight values in the threats data table which is input into habitat quality model.
To calculate the habitat degradation caused by urban areas, the formula could be:
access_layer.tif x relative weight_urban x sens_urban.tif x filtered_urban.tif.
Besides, may I know what does “LandUse_aligned.tif” in the intermediate folder represent? Does it equal to the input LandUse map?
Does habitat.tif in the intermediate folder equal to Hj in the following screenshot? Many thanks!
access_layer.tif, sens_.tif and filtered_.tif are all in the intermediate folder.
Yes, these files are in the intermediate folder.
Relative weight could be calculated by myself from weight values in the threats data table which is input into habitat quality model.
Yes, you would sum all the weights in the table column and then for your single threat would normalize by <threat weight>/sum_of_weights. That is what you would use.
To calculate the habitat degradation caused by urban areas, the formula could be:
access_layer.tif x relative weight_urban x sens_urban.tif x filtered_urban.tif.
Yes, that’s right.
may I know what does “LandUse_aligned.tif” in the intermediate folder represent?
The model takes all the input rasters and makes sure that the pixels align to do pixel operations. So yes, this is just the input land use raster.
Does habitat.tif in the intermediate folder equal to Hj in the following screenshot?
Yes, that’s right. This is the value found in the sensitivity table under the Habitat column.
Unluckily, I found it was above the ability of my laptop to multiply rasters directly since these rasters were large. So I have tried to run Habitat Quality Model with each threat.
For example, in order to calculate degradation caused by urban areas, I only retain the threats data of urban areas and habitat sensitivity to urban areas. That is to say, I run habitat quality model which only considers urban threats. Then I multiply the habitat degradation from the model by relative weight of urban areas.
In this way, I think I could get degradation values caused by each threat —— and the sum of these values equals to the total degradation value which is obtained by running the model with all threats.
Based on the formulas of habitat degradation calculation and previous discussion under this topic, I believe this method is logistically right. May I know your opinion on the above method? Many thanks!!
is the relative weight of the threat
, the distance-weighted threat raster, is 