Coastline Population Density Analysis

Question

Original User: BeaconEnviro

Hi Everyone,

I am trying to create a raster base on a series of population measurements along a coastline. I would like to create an output similar to this:

http://soundwaves.usgs.gov/2009/03/MapLinearDensityLG.jpg

My data consists of a series of points spread along a couple hundred kilometers of shoreline, each point has a number assigned to it based on population evidence (# of burrows). The points are at non-regular intervals and consists of wide areas with no data and a large amount of clumping at other locations.

I have tried several methods in both Spatial and 3D analyst, but I have yet to come up with a single raster that is not empty. Is there a technique that will work on data in a linear formation?

Thanks for any help you can give.

After.gif

Before.gif

2025-11-13_09-45-03.png

JeffreyThompson2_1-1763055143515.png

WilliamHuber · Answer

Is there a technique that will work on data in a linear formation?

By specifying "linear" this question was asked in exactly the right way, because almost any attempt to do the analysis in two dimensions will produce biased results.

Instead, represent the locations as measured distances along the coast, perform a kernel smooth of those points in the single (measure) dimension, and transfer the smooth back to the measured coastline.

The kernel smooth can be done in a myriad ways with software ranging from Excel to ArcGIS to R, depending on what you're comfortable with.  Transferring it back to the coastline is not so easy.  One way is to dissect the coastline into segments, compute the average smoothed value within each segment, and attribute those segments with the averages.  This can then be displayed as color (or pattern) coded line segments on the map.  This gives you much better cartographic control than a 2D grid, anyway.

The example offered is interesting for the cartographic distortions it introduces: by varying the apparent width of the colored coast, it alternately emphasizes and de-emphasizes the sea otter population.  This width does not seem to convey any information of its own--maybe it's just a map-making artifact.

Anonymous User · Answer

Original User: SaultDon

You could buffer your coastline so that you make a study area like in the image shown.

Then create lines that intersect the coastline where points would lay if they were exactly on the coast line. So your finished result will be something that looks like a railroad track. (the intersecting lines should be at 90deg angles where they meet the coastline centreline)

Attach the point values to their corresponding intersecting line.

Now what you could do is run these lines to create either a TIN or a more gradient resulted option would be to produce a DEM with the coastline buffer as the extent using the z-values from the intersecting lines.

Or you could just use eucl. dist. on each point individually, then sum the cells. If using this method, attention will need to be used as you will have a No Data impact when summing the cells if you do not use the con function to set the null values to an integer >= 0
I would suggest building a model for this and exporting to python for easier execution.