Interpreting outputs in corridors.gdb

[Arne Loth]

Hi,

I've just run the 'Build Network and Map Linkages' analysis in ArcMap 10.6 using Linkage Mapper 1.1.1. I previously posted about some trouble I was having with this as my LCP file extended beyond my study area and produced a 'lattice' effect. This issue has now been resolved.

Looking at the output, I am somewhat confused about the normalized and mosaicked least-cost corridors found in the corridors.gdb. In my analysis I chose a Maximum Euclidian Corridor Distance of 2000 meters under 'Additional Options'. I thought that this would effectively limit the corridors file to a buffer of 2km around each pair of core areas that are connected. However, what I ended up with was a file called 'corridors_truncated_at_200k'. What does '200k' refer to? Surely this cannot mean 200,000 meters Euclidian distance, so is in cost-weighted distance units?  If so, what effect did the 'Maximum Euclidian Corridor Distance' setting have, i.e. what is the relationship between 2000 meters maximum Euclidian distance and the 200k in my corridor output? 

I need to be sure that I fully understand these outputs as I am supposed to explain my maps (see example below) to people with no background in spatial modelling so any advice would be greatly appreciated. 

Many thanks,

Arne

[John Gallo]

Hello Arne,

You are almost there! You did not adjust the width, you set the maximum euclidian distance of the least cost path between two cores.  So, if two cores are more than 2000 m apart, the linkage does not get mapped.

Regarding the truncated output, that is a default truncation of 200,000 cost distance units.  (The width of your linkage.)  Want to change that default?

Here is a key part from the manual:

"Once you are familiar with Linkage Mapper, you can change some settings, such default corridor widths or whether to use minimum distances between core areas, by editing lm_settings.py (located in your toolbox/scripts directory)."

note, you can also change the width of the non-truncated output, either manually or with the Utilities.

If I remember correctly, we made this more clear int he version 2.0 documentation, but I am not positive.  As I've mentioend elsewhere on this list if you would like to beta test 2.0 just let me know and I'll send it along.

Thanks,

John

[Arne Loth]

Hi John,

Many thanks for your quick reply. This makes more sense now - I did notice that my least-cost paths were capped at 2000 m. Is the cost-weighted distance space measured in map units, i.e. 200,000 meters?

Cheers,

Arne

[John Gallo]

Hi Arne,

Not quite.  It is essentially Map units X cost.  Just be careful about how many map units are in a cell.

From Esri regarding cost weighted distance:

"The cost assigned to each cell represents the cost per unit distance for moving through the cell. The final value per cell is the cell size multiplied by the cost value. For example, if the cost raster has a cell size of 30, and a particular cell has a cost value of 10, the final cost of that cell is 300 units." 

So it would take a corridor width of 10 such cells to meet a CWD threshold of 3,000.

Cheers!

John