GIS 2001

Lab G - Vector Overlay in NVMD

VECTOR OVERLAY

This exercise is designed to examine the relationships between 1940 land use data and 1988
land use data by performing overlay operations using the appropriate GIS layers. We will be
using the Analysis Tools in Arc ToolBox to perform the overlay operations. The purpose of this
lab is to look at the overlay operations available in ArcGIS, learn how they work, how they
create new GIS layers with attribute tables from the combination of the input layers, and to
answer questions using vector overlay.

The Data

The data for this lab is available on the I:\Drive. Download the Lab G folder; It consists of 2
shapefiles: luse40.shp and luse88.shp. In addition, there is a text file named landuse_codes.txt.
The polygon shapefiles represent land use categories in 1940 and 1988, which are coded in the
attribute table. The text file can be used to decode the land use codes.

The Study Area

This data comes from the Northern Virginia Military District, an area in what is now Ohio that
was set aside historically by the state of Virginia as payment to Veterans of the Revolutionary
War. More information about this District can be found here:

http://en.wikipedia.org/wiki/Virginia_Military_District

Here it is in GoogleEarth:

Lab G – Vector Overlay Page 1 of 7

Figures 1a and 1b. Study area in GoogleEarth

Geologically, the study area can be divided into an undulating northern till plain and a very flat
southern outwash plain, both parts of glacial geologic systems, bisected by Darby Creek, a
riparian (river) area.

The Questions

For this lab, our activities will be directed toward examining how land uses have changed in this
area over time. Land use refers to human activities (or lack thereof) on the landscape.
Historically, the NVMD has been an active agricultural area, but increasing urbanization has
impacted the historic human landscape, as well as the forests in the area.

Specifically, we want to examine the northern and southern parts of our study area separately to
see how changing land use has impacted traditional agricultural lands and forests. In terms of
urban expansion, has it occurred on agricultural land from 1940 to 1988 in the till plain (northern
portion of the study area)? Has it occurred on agricultural land from 1940 to 1988 in the
outwash plain (southern portion of the study area)? What can we find out about the process of
urban encroachment on agricultural land on these areas from 1940 to 1988? (That is, did urban
encroachment occur around existing urban areas and/or did new urban areas appear over this
time period?)

In relation to the forests, we would like to know how much change in the amount of young
woodland to upland forest has occurred over the entire study area from 1940 to 1988. Where is
this change most notable (outwash plain or till plain)? Finally, we would normally expect to see
young woodland continue to change to upland forest. Where does it not?

The answers to these questions can all be derived from the GIS layers and data table that you
have been supplied with for this exercise.

Because we are using 2 distinct GIS layers representing the same areas in different periods of
time, overlay operations will be extremely useful to us. Overlay operations in GIS enable us to
take 2 or more spatially coincident layers and perform operations on them. Before you begin
working on this lab, take some time to review the following overlay operations presented
in the literature and other exercises, and in using the Help documentation in ArcMap:
Intersect tool/command, Union tool/command, Identity tool/command. Where are these
tools located in the ArcToolbox structure?

Lab G – Vector Overlay Page 2 of 7

Data Preparation: Create the Geodatabase

Now we need to create the geodatabase and feature dataset. Before we can create the feature
dataset we need the spatial reference information for these shapefiles. Examine the properties
of the luse88.shp and the luse40.shp shapefiles. What is their spatial reference? It is
undefined! You can look at the properties for both of these shapefiles until you want to cry and
at best find only some information on bounding coordinates. Write down the bounding
coordinates for the luse40 shapefile, noting whether the values belong to left, right, top, and
bottom. The values should be similar to:

Left: 287329.875000
Right: 297787.687500
Top: 4462131.500000
Bottom: 4438203.500000

What are these values? They are the values that make up the coordinate pairs that define the
diagonal corners of a rectangle containing the entire study area. That is, the 2 coordinate pairs
are also known as the extents of the study area.

How could you find the correct spatial reference? You could examine the values in the
bounding coordinates and combine that along with the knowledge that the study area is in Ohio
in order to guess the coordinate system, although you would not really know the correct datum.
You could look for the Northern Virginia Military District on the Internet and find out the study
area location, hopefully get some GIS layer(s) with known projection and datum information that
cover all or part of the study area, and figure out this information for the study area GIS layers.
This might take some concerted effort on your part. On the other hand, if you received these
layers directly from someone, it would be better to contact the person and request the
information directly from them. But being that this is a class exercise and time is short, the
projection and datum information is: UTM Zone 17N NAD27, meters (old school!).

Before we export these shapefiles to a geodatabase, we will define a spatial reference for each.
If you do not do this, you can run into odd issues related to the XY extents of your data later on.
Since we now have the correct spatial reference information, we have all of the information
required to define the spatial references. Do you recall how to do this? Do it!

Next we need to first create a geodatabase. In ArcCatalog navigate to the location where you
want to store your data and select New  File Geodatabase. You can rename the
geodatabase to something informative like “NVMD”.

We now need to create the geodatabase feature classes Geodatabase from the 2 shapefiles
luse40 and luse88. In the ArcCatalog window select the luse40 shapefile and right-click. Move
the cursor to Export and select To Geodatabase (single)… In the Feature Class to Feature
Class window, the name should already be filled in. In the Output Location dialog window,
browse to and double-click on the NVMD geodatabase icon and click Add. In the Output
Feature Class Name dialog window, type in landuse1940, then click OK. Do the same for the
luse88 shapefile, naming the feature class landuse1988.

If you expand the Geodatabase in the ArcCatalog Table of Contents, you should now see the 2
new polygon feature classes, “landuse1940” and “landuse1988” (mine are named luse40 and
luse88, below and there are additional feature classes shown below as well).

Lab G – Vector Overlay Page 3 of 7

Figure 2. The NVMD Geodatabase

Examining the attribute tables for these layers, you will also notice that there are 2 fields each
for Area (Shape_Area) and Perimeter (Shape_Length). Why is this? I’ll give you a hint: when
you create a feature class, it automatically calculates an area and length and puts them into
fields named Shape_Area and Shape_Length. These are the terms for the attributes of the
shapes, and are used for other calculations.

Data Preparation: Join a table to the GIS data

Examine the attribute table for the 1940 layer. One of the fields you see is LUSE_CODE. This
field contains the numeric codes for the landuse categories. By themselves, these codes are
not very useful as nothing comes to mind when you think of landuse code #206. We need a
codebook to translate these codes into something more useful. Using Windows Explorer, you
will notice a file called Landcode.txt. This is a text file, and if you open it you will see the
description for each of the landuse codes. Since we will be selecting specific landuse codes,
you will need to refer to this information. There are 2 ways to do this. The first is to refer to the
information in the text document by either printing it out or accessing it online. The preferred
way is to Join the information in the text file to each of the feature classes in the feature dataset.
To do this you would need to create a table of the information that can be joined to the feature
attribute table. Specifically, you would need to store the information in the text file as a data
table in the NVMD geodatabase.

You may have a more efficient strategy, but since this is a small table, here is what I did:

 Open landcode.txt in Excel, import as a fixed width file

 Add a row at the top of the table
 Name column A: LUSE_CODE
 Name column B: COVERTYPE
 Save the excel spreadsheet as landcode.xlsx
 Browse to the NVMD Geodatabase in ArcCatalog
 In the Geodatabase, right-click and select Import Table, then OK
 In the Import window, browse to the landcode.xlsx file.
 Rename the table in the NVMD Access database to landcodes

Lab G – Vector Overlay Page 4 of 7

You should now have a Geodatabase that looks something like the one shown below:

Figure 3. The NVMD Geodatabase with table

Note: if the table fails to import into the Geodatabase, just use the .xlsx spreadsheet to do the
rest of the lab.

Start ArcMap and add the landuse1940 and landuse1988 feature classes. Also add the
landcodes table to ArcMap. Now, join the landcodes table to both of the feature classes. You
need to choose the correct field to do this. You want to end up with the description of the land
cover in the table (e.g., Young Woodland, Urban, etc.).

As you did in your Maine lab exercise, to permanently add the landuse code descriptions from
the landcodes table to the feature attribute table, we need to create a new field and calculate
the values for this field. Click on Options in the feature attribute table and click on Add Field.
Make the name of the new field COVERTYPE40 for the 1940 landuse layer and
COVERTYPE88 for the 1988 layer. For the type of field to create, choose Text and make the
width of this field 30 characters. Click OK. Find the new field in the attribute table and right-
click on the field name. Select Field Calculator and then click on Yes in the next window that
appears. We want to calculate the values found in the field within the Field Calculator window.
See if you can figure out how to do this.

After this has successfully executed, right-click on the landuse1940 feature class, select Joins
and Relates… and then select Remove Join(s). In the feature attribute table for the layer you
will now see that the fields from the Landcodes data table have disappeared and the name of
the feature class appearing before the fields in the feature attribute table have also disappeared.
The data from the Landcode data table are now permanently stored in the feature attribute table
for this layer in the field covertype40. Perform the same sequence of operations for the
landuse1988 feature class.

Lab G – Vector Overlay Page 5 of 7

Figure 4. Attribute table with permanently added vegetation descriptions

Perform the Analysis

The overlay tools for feature classes can be found in ArcToolBox. Expand the Analysis Tools
and then expand the Overlay toolbox. You will use one of the types of overlay operations
reviewed in the materials. If you choose to perform an Intersect overlay, for instance, double-
click on Intersect and fill in the dialog windows that appear in the Intersect window. Leave the
Join Attributes setting as ALL, the Cluster Tolerance dialog window empty and the Output Type
dialog window as INPUT. This last setting will create the same type of feature class type as the
input feature class types, such as polygon. After an overlay operation, open the feature
attribute table of the Output feature class you created and examine the table for the fields that
are present in this new feature class.

You will use attribute selections (SelectionSelect by Attribute) to answer the questions. You
may want to export the selected polygons to new layers in order to keep things straight. You will
want to view the statistics summaries as well.

Please be aware that intersecting layers results in ALL the attributes from both layers
appearing in the output layer. You need to be careful that you are looking at the right
attributes. It may be helpful to delete unnecessary attributes after performing an
intersect operation!!!

Hand in responses to the following review and discussion questions for this exercise:

1. Has urban expansion occurred on agricultural land from 1940 to 1988 in the till plain
(northern portion of the study area)? Compute and include any change in area that
occurred under this condition.

Which type of overlay did you use to answer this question, and why?

Create a map of the resulting overlay to turn in. Clearly indicate the polygons that fall

Lab G – Vector Overlay Page 6 of 7

 into the class of urban expansion on agriculture from 1940 to 1988.

What does this tell us about the process of urban encroachment on agricultural land on
the northern portion of this study area from 1940 to 1988? (That is, did urban
encroachment occur around existing urban areas and/or did new urban areas appear
over this time period?)

2. Has urban expansion occurred on agricultural land from 1940 to 1988 in the outwash
plain (southern portion of the study area)? Compute and include any change in area
that occurred under this condition.

Which type of overlay did you use, and why?

Create a map of the resulting overlay to turn in. Clearly indicate the polygons that fall
into the class of urban expansion on agriculture from 1940 to 1988.

What does this tell us about the process of urban encroachment on agricultural land on
the southern portion of this study area from 1940 to 1988? (That is, did urban
encroachment occur around existing urban areas and/or did new urban areas appear
over this time period?)

3. How much change (in area, and state the map units) in the amount of young woodland
to upland forest has occurred over the entire study area from 1940 to 1988?

Where is this change most notable (outwash plain or till plain)?

Create a map of the resulting overlay to turn in. Clearly indicate the polygons that fall
into the class of young woodland changing to upland forest from 1940 to 1988.

4. We would normally expect to see young woodland continue to change to upland forest.

Where does it not?

That is, compute the total area where young woodland changes to another land use type
other than upland forest.

Create a map depicting this change and showing the land use type that young woodland
changed to, other than upland forest.

Lab G – Vector Overlay Page 7 of 7