Kernel Interpolation with Barriers -- barrier restrictions and algorithm used

4344
2
04-20-2015 12:13 PM
KrystanWilkinson
New Contributor

Hello,

I am interested in learning more about the Kernel Interpolation with Barriers tool provided by the ArcGIS 10.x software.  Upon review of the ESRI help files and webpage discussions, I am left with a few questions regarding this tool and was hoping someone could provide some insight:

1.       Are there restrictions as to the complexity of the barrier?  For example, I have used the adeHabitatHR package in Program R which uses an algorithm suggested by Benhamou and Cornelis (2010), but the input boundary has 2 constraints:

    1. “The boundary should be defined as the union of several segments, and each segment length should at least be equal to 3xh, where h is the smoothing parameter used for kernel smoothing;
    2. The angle between two successive line segments should be greater than π/2 or lower than –π/2”

It is unclear from the help files if the Kernel Interpolation with Barriers tool in ArcGIS has similar requirements for the input barrier, or if all barrier shapefiles are acceptable.  I am working with a fairly complex coastline.

2.       The Kernel Interpolation with Barriers help file indicates: “The absolute feature barrier employs a non-Euclidean distance approach rather than a line-of-sight approach.”  Which non-Euclidean distance approach does the barrier feature utilize?

Much thanks in advance!

Krystan

0 Kudos
2 Replies
EricKrause
Esri Regular Contributor

Hi Krystan,

The distance metric that is used is the shortest sequence of straight lines that go around the barrier feature class.  This is what is meant by "non-Euclidean."

There are no limitations for the geometry of the barrier feature class, but the more complicated the geometry, the longer the method will take to process.  If you are experiencing performance issues, you can use the Simplify Line or Simplify Polygon geoprocessing tools to remove vertices from the barrier feature class, and this will dramatically speed up the calculation.

You can read the full details of the algorithm in this paper:

http://link.springer.com/article/10.1007%2Fs00477-011-0488-2

KrystanWilkinson
New Contributor

Dear Eric,

Sincere thanks for your response.  It was very helpful.  Thanks for the link to the paper as well!

Krystan

0 Kudos