difference between the generated CWD results and my original resistance surface？

[廖春在]

Hi everyone! I think I’m running into a non-technical issue. My resistance surface ranges from 0.08 to 4.95. When I run the linkage analysis (the first step, Build Network and Map Linkages, with a Truncate Cost-Weighted Distance Threshold of 500), I notice that the generated CWD values are much higher than my original resistance surface range. Is this normal? I also found that the CWD values in the linkage_LCPs results are significantly higher than what my resistance surface would suggest. For example, one corridor has（on the table ，2 line）: cwd_dist = 3767, euc_dist = 1524, lcp_length = 1661, and cw_to_Euc_Dist_Ratio = 2.47.

I calculated the number of pixels along the path: lcp_length / pixel size = 1661 m / 30 m/pixel ≈ 55.4 pixels.
Then, the average pixel resistance is: cwd_dist / number of pixels = 3767 / 55.4 ≈ 68.0.
This average pixel resistance of 68.0 is much higher than the maximum value of my resistance raster, which is only 4.95. I’m wondering if the path here corresponds to the raster of the LCP line segments, or does it represent a range, like in the corridor TIFF?

I’m eager to understand where I might be going wrong and would greatly appreciate any guidance! Thanks so much!

[John Gallo]

Hello Lia Chun Zai,

You are on the right path!  (ha ha, pun intended).

But cost weighted distance is in map units not pixels.  try cwd_dist / (number of pixels * number of map units per pixel edge)

In this case map units are meters and number of map units per pixel edge is 30.

Did that work?

Also, an emerging best practice is to have the lowest cost value on the landscape = 1, and then all costs to go up from there.  Then movement in this space is = to the energetic cost of movement only, and is equal to distance.  Not mandatory, but a nice touch to consider adding.

Good luck!

John

--
You received this message because you are subscribed to the Google Groups "Linkage Mapper" group.
To unsubscribe from this group and stop receiving emails from it, send an email to linkage-mappe...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/linkage-mapper/6f4fab70-e7ed-44ac-a12b-8dfcc999108en%40googlegroups.com.

--

John A. Gallo, Ph.D.

Project Director

LinkageMapper.org

and

Senior Conservation Scientist

Conservation Biology Institute

[liao mike]

Your prompt reply and scientifically spot-on explanation have left me in awe! I think I’ve finally cracked the case: my blunder was overlooking the geographic scale of the pixels themselves. It turns out, the correct way to interpret cwd is by factoring in the pixel’s geographic size multiplied by its corresponding weight. Moving forward, I’m excited to experiment with setting the minimum resistance value to 1 to polish my results further.
A huge thanks to you and the Linkage Mapper team for your patient, timely, and precise guidance! While this might just be a “rookie” mistake on my part—perhaps the kind that deserves a gentle facepalm—I’m beyond grateful for your kind and thorough response.

 
Wishing you all the best in your work,
Liao Chun Zai

John Gallo <john....@linkagemapper.org> 于 2025年4月17日周四 01:45写道：

To view this discussion visit https://groups.google.com/d/msgid/linkage-mapper/CAEUWjEvWTLo5Mb%3DfrBuo8y%2Bq5Xfrv4QN%2BXZnhtDxxvEHgGtVLg%40mail.gmail.com.

[John Gallo]

Hi Liao Chun Zai,

Thank you for the very kind email.  Provides fuel for this work.  And no, not a rookie mistake.  Most rookies never evben take the time to figure out what the CWD numbers actually mean.

On a separate note, thanks for sharing the resistance surface and initial outputs.  I'll send you a private email with a couple quick suggestions on possible improvements for connectivity conservation, depending in part on the species or ecological process you are modeling.

Thanks,

John

To view this discussion visit https://groups.google.com/d/msgid/linkage-mapper/CAOG9Zt77J59rMAc2Kt-8jG2t-0VRbcBGAAZX9kpEq21s55h8XQ%40mail.gmail.com.