Skip to end of metadata
Go to start of metadata

ArcMap provides several different tools for proximity analysis. From the System Toolboxes, open Analysis Tools, and click on Proximity.

Tools available under Proximity are:

  • Near, Generate Near Table, and Point Distance (all under three separate headings)
  • Buffer and Multiple Ring Buffer (under separate headings)
  • Create Thiessen Polygons

Near

The various flavors of Near, Generate Near Table, and Point Distance are all very useful for finding out what things are closest to other things. For a set of sites, where one wants to find the closest site and know that distance, the Near tool can be run on a single featureclass. For each site in the feature class, the tool will find the closest site (that is not the site itself). The OID of the nearest site and the distance to that site are appended to the attribute table of the feature class. The fields are NEAR_FID and NEAR_DIST respectively. The tool can also provide bearing. For most archaeological purposes, knowing the site and that distance is a useful beginning. However, many different combinations of points and polygons can be implemented. In this instance, we will work with a simple set of points.

  1. Add the sites feature class.
  2. Click on the search tool, type near into the search box, and hit return.
  3. Select the Near (Analysis) option. In the dialogue box, for the Input Features select the sites feature class. In the same dialogue box, for the Near Features, select the same feature class. For this exercise, there is no need to restrict the search radius nor report the Location and Angle, but these are options to bear in mind for other applications. For now, just accept the default outputs and select OK.
  4. Check the attribute table and see if there are two new fields named "NEAR_FID" and "NEAR_DIST".
  5.  Now click on the NEAR_FID and sort this column in either ascending or descending order.
  6. The field NEAR_FID identifies what site is closest to the site that corresponds to the given row. Are some sites close to more than one other site? In terms of centrality, what might this imply?
  7. Now check the "NEAR_DIST" field. Right click on this field, select Calculate Statistics, and observe the result. What are the minimum, maximum, mean, and standard deviation distances? Given your inspection of the distribution in the reported histogram, do you think that the mean is an appropriate measure of the central tendency of this distribution? Just looking at the table may not provide enough information. At this point, it would be useful to explore the data in a spread sheet or statistical package.
  8. There are multiple ways to get data from a table.
    1. The simplest method is to open the feature class attribute table, click on the table icon in the upper left corner of the window, select "Export" an direct the dialogue box to create a text or dbf file in the desired location.
    2. Another method is to select all rows and right click on the grey area on the left side of the table, select copy, and paste these cells into the desired application. This web page contains some useful information on copy and paste of rows into another application.
    3. A third way to obtain the nearest feature data in table form is to use the Generate Near Table tool. Under Proximity, select the Generate Near Table Tool, indicate the proper Input and Near Features (following guidelines as described above), provide the Output Table location and name, and select OK.
  9. Once the table is exported, open it in an application like Excel or PAST. Compare the mean, median, and mode. Based on your understanding of the distribution, what is a good measure of central tendency for this sample?

Thiessen Polygons

These features can be useful for examining neighborhoods around a site or some other feature. Considerable caution may be required in interpreting the results, especially when topographic variation or other geographic constraints dictate or influence a distribution. Still, Thiessen polygons can help illustrate interesting structure in a distribution of features. The method has seen application in archaeology, and merits consideration.

Thiessen Polygons are created by drawing lines between a set of points, taking the midpoints of these lines, and building polygons from those derived lines. Where sites are bunched together, Thiessen polygons are smaller. Where sites are dispersed, Thiessen polygons are large.


Thiessen Polygons around a set of sites.


Annotations showing how the Thiessen Polygons are created.

Create Thiessen Polygons

  1. If it has not already been done, add the selected feature class for which the Thiessen polygons will be constructed.
  2. Under Catalog, navigate to Toolboxes>Analysis Tools>Proximity>Create Thiessen Polygons.
  3. Select the input point feature
  4. Define the location and name of the output feature
  5. Define an output field if desired. Site name, site number, or some other identification is important for linking back to the input features.
  6. Select OK to generate the Thiessen Polygons.

Now take a look at the output.

Often times once the Thiessen Polygons are created it is necessary to cut them and frequently one wants to make these cuts via another polygon. In this case, we created Thiessen polygons along the coastline of Peru. If one is comparing terrestiral neighborhoods, it would be necessary to clip the Thiessen Polygons based on the coastline. Likewise, if one is comparing terrestrial neighborhoods within valleys, it makes sense to clip the Thiessen Polygons along watershed boundaries.

There are several ways to clip polygons by the boundaries of other polygons.

  • Trace Feature Option. This ESRI blog post has some information on the trace feature option. This can work well when there are few intersecting polylines or polygons. However, when there is complex overlapping geometry, the tracing tool can prove problematic.
  • Union Tool Option. In more complex settings like those just described, it is advisable to employ the Union Tool. This tool can be found at System Toolboxes>Analysis Tools>Overlay>Union. While Union does a fine job slicing up complex polygons, the tool can make multipart features and these can be problematic when trimming out unwanted sections of the result from the Union operation. Intersect is another potentially useful option for complex clipping tasks. Intersect also creates multipart polygons. To explode multipart polygons, bring up the advanced editing tools and select the explore multipart option. Once the multipart polygons are exploded, select and delete the individual unwanted polygons.

Clip Thiessen Polygons and clean up unwanted features

  1. If it has not already been done, add the Thiessen Polygon layer.
  2. Add the Watersheds_SRTM polygon layer.
  3. In the System Toolbox navigate to Analysis Tools>Overlay>Union.
  4. Select the Input Features, define the location and name of the output feature, and click OK.
  5. Examine the resulting feature class. Some of the unwanted polygons should be removed. Clicking on them shows that several of the polygons are multipart and some of the parts are wanted. Given this, it is necessary to explode multipart polygons.
  6. Select the output from the Union operation and begin an edit session.
  7. From the Edit menus, select More Editing Tools>Advanced Editing
  8. Under the Advanced Editing toolbar, click on the Explode Multipart icon.
  9. Save the Thiessen Polygon layer.
  10. Select and delete the unwanted polygons, save the layer again, and end the edit session.
  • No labels