Massachusetts Institute of Technology
Department of Urban Studies and Planning

11.520: A Workshop on Geographic Information Systems
11.188: Urban Planning and Social Science Laboratory

Lab Exercise 2: 
Thematic Mapping in ArcMap

[Click here for today's in-class notes. ]

In this exercise, you will build on the basic ArcMap techniques you explored in Lab 1 to make more sophisticated thematic maps. You will make two different kinds of maps: exploratory and explanatory. In the exploratory section, you will compare two thematic maps of the Cambridge population. One map shows population counts by block group and the other normalizes the population counts by dividing by land area to show population density. In the explanatory section, you examine the spatial pattern of home sales prices to see if they appear to be related to housing density. Finally, you will convert some of your finished maps into Portable Document Format (PDF) files, a format that facilitates electronic distribution of your work.

I. Setting Up a Work Environment

Step 1. Attach 11.520 Locker. This is exactly the same step as in Lab1.

Much of our class data reside in the course locker, 11.520. This locker is accessible via the Andrew File System (AFS) as Z:\afs\athena.mit.edu\course\11\11.520. However, ArcMAP can have trouble trying to read upper-level directories on AFS drives. (ArcMAP insists on trying to find mapable data in each sub-directory that it examines. The upper-level directories on AFS are spready among machines around the world and ArcMAP will spend forever looking.) Since we will use the class locker repeatedly, we might as well map the locker to a 'virtual' local drive and avoid this problem. To map the class locker to your M:\ driver, open a DOS command window, click Start and choose "Run.... "

In the DOS command window, type: "attach -Dm 11.520" and press enter. This will mount //afs/athena.mit.edu/course/11/11.520/ to drive M:\ for access to the class locker 11.520. Later we will add data from this locker.

Step 2. Start ArcMap by follow the following steps.

1. Click the Start button in the left bottom corner.
2. Move the mouse over the item Programs.
3. Move the mouse over the item ArcGIS
4. Move the mouse over the item ArcMap and Click.

Please wait patiently for ArcMap to launch. The program takes awhile to come up.

II. Effective Mapmaking

As in Lab 1, we will use 1990 U.S. Census data for Cambridge for this exercise. In ArcMap, add cambbgrp (the Cambridge Block Group shapefile in the M:\data directory). Open the Data Frame Properties (View > Data Frame Properties from the menu bar) window and set the "Map Units" to meters and the "Display Units" to feet.

A. Exploratory Mapping

Exploratory mapping is the kind of mapping we do to get a feel for what the data "looks like." Before coming to any conclusions, you should make a number of maps that summarize what is in the database and raise questions that you would like to answer. We will go through a number of exploratory mapping techniques here.

The Effect of Normalization on Thematic Maps

Using the methods described in Lab 1, use the 'symbology' tab of the layer properties to modify the characteristics of the cambbgrp theme to match these criteria:

Use the Layer Properties window to change the name of this theme to Population. To change the number of decimal places, put the cursor on a label, click right mouse butten. It will open a Number Format window. Since numbers represent population, choose numeric and set the number of decimal places you want. In this case, set 1 for number of decimal degree.

Fig. 5. Open a Number Format window


Fig. 6. Set the number of decimal places

You have created an unnormalized map of Cambridge population by block group.  Your map isn't as meaningful as you might like since your thematic shading depends only on the raw number of people in each block group - regardless of whether a block group is itself large or small.  In such cases where your raw data are not 'normalized' (i.e., adjusted to represent a compared-to-something-meaningful  comparison), it is easy to generate a pretty looking map that is quite misleading.  Shortly, we'll compare this thematic map with one that does normalize the population count.  But, first, lets spruce up the map a bit.

Click on the Add layer button , and add majmhda1.shp from M:\data . This is a MassGIS (Massachusetts Geographic Information Systems) Major Roads Datalayer created in December 2000.

Now let's adjust the characteristics of the major roads theme. First, change the name of the layer to Major Roads. Next, open the Layer Properties window and click the symbology tab and set the properties as follows:

This is the first time we have encountered a "Unique Value" legend type. This type of legend is often useful when a field takes on only a limited number of discrete values. In this case, the field Class assumes only four values: 1, 2, 3, and 4. Each value represents a different type of road. These numbers are not very descriptive, so we will want to include text labels for each value. But what do the numbers mean? For that, we must consult the MassGIS metadata for this dataset. Use the descriptions for the values of Class there for the labels in your legend.

By default, ArcMap does not choose a very attractive symbolization scheme for the roads. You will need to adjust the symbols manually. Set the symbols as described in Table 1.
Value Color Size Style
1 
 Red
3 
 Solid
2 
 Blue
2 
 Solid
3 Dark Gray
0.1 
 Solid
4 
 Dark Gray
0.1 
Dotted
Table 1. Road Symbolization

 

 

 

 

 

 

When you're done, your layer properties-symbology window should resemble Fig. 7. Note that the full text of the labels is cut off in Fig. 7; you should enter the full text of the description as documented in the dataset's metadata. Don't forget to apply your changes!

Now you will create a normalized map of the population data. Normalizing means adjusting for effects that distort the way the data appears. Here, you will compensate for the effect of land area on population. Larger areas typically have larger populations than smaller ones, but the actual densities of the areas may be quite different.

Unique Value Legend Window
Fig. 7. Unique Value Legend Window

Create a new data frame by select insert "Data Frame" menu from the menu bar. You should now see the "New Data Frame" icon in the data frame window. Select the Major Roads layer under "Layers", click the right mouse button and select "copy". Then select the "New Data Frame", click the right mouse button and select "Paste Layer". Now paste the Population layer into the "New Data Frame". Remember to set the Map and Display units for this new data frame.
Insert a New Frame New data frame appears Copy layers from existing frame Paste layers to the new frame
Fig. 8. Create a New Data Frame


Open the layer properties to modify the Population layer.  Adjust the number of decimal places to "1" and set the "Normalize by:" field to "Landacre." In ArcMap, the 'normalize by' option is used to pick an attribute that will be divided into the mapped attribute before doing the classification and shading.  Hence, normalizing by landacre will create a population per acre measure. You should still be using the "Quantile" classification with 5 classes and "Orange Monochrome" color ramps. When you apply your changes, you will have a population density map. Change the layer's name to Population Density. Mathematically speaking, what did the setting the normalization field do to the population values?  Is the population density map more consistent with your impression of the parts of Cambridge that are more crowded?  Also, compare the result when you normalize by the 'Area' attribute rather than the 'Landacre' attribute.  The 'Area' attribute is the area (in square meters) of each block group polygon whereas the 'landacre' measure excludes bodies of water (like the Charles River Basin or Fresh Pond).

Since you have two data frame and data view displays only one frame at a time, you need to activate a frame to see it. For example, if you want to see "New Data Frame", click right mouse button with the mouse over "New Data Frame", and then select "Activate". (Note, you *can* show two maps within one "layout" view. We will cover that soon.)

 Fig. 9. Activate a Frame


Annotating Your Map

While a map often can tell a convincing story on its own, you can often strengthen your case with well-chosen labels and annotation.

Highlighting an Area of Interest

You should be able to spot one or more areas in your two maps of Cambridge where the discrepancy between the two maps is especially apparent. In the original "Layers" frame (the unnormalized map), highlight one of these areas by drawing a circle around the area with the New Circle tool Draw Circle tool. To select the New Circle tool, you will first need to open the draw tool pop up menu as shown in table 2 and select the New Circle from the pop up list of icons that appears. The complete list of drawing tools is shown in Table 2.

Drawing tools New Marker
New Line 
New Curve
New Rectangle
New Circle 
New Polygon 
New Ellipse
New Freehand
Table 2. Drawing Tools

Once you've drawn the circle, you can use the Arrow Pointer tool Arrow Pointer tool  to move and resize it. You can change attributes such as the line thickness, outline color, fill pattern, and so on using the Properties Window. To open the properties window, double click the circle you just created. You can also open the properties window by clicking the right mouse button while placing the cursor over the circle, and then selecting properties from the botton of the drop down list that appears. Make your circle have an outline width of 3 and change the color to something that stands out, yet doesn't clash with the color scheme of the map.

Adding Text Annotation

Now use the "A" Text tool Text tool to add some text annotation near the circle explaining why you put it there (e.g., "Zone of High Discrepancy"). Use only a few words and choose font characteristics that will make the text visible but not overwhelming. Note that, as with the drawing tools, you can choose an annotation style from a pop up list. You can experiment with these if you wish.

Adding Labels

Labelling features on the map can help viewers orient themselves. Labelling too many features, however, clutters up the map and impedes readability. Let's put identifying markers on stretches of limited access highways, symbolized by thick red lines on your map. Use the "i" Info tool Info tool to look at the attributes of the highways. Make sure that the Major Roads layer is selected first. You will probably find that at least two records show up for each click; that's because the two directions of the multilane highways have been encoded separately, as if they were two different routes very close to one another. If you watch closely as ArcMap flashes the matching links, you can actually see this.

Interstate 90 (a.k.a. the Mass. Pike) extends east-west near the bottom of the image. As you identify links along this stretch, you should see that the "Rt-number" field is "90" and that the "Admin_type" is "1." From the metadata, you can see that an Admin_type of 1 indicates an Interstate highway. We want to label a few roads using their "Rt-number" field. Open the Layer Properties window for this theme. Click on the "Labels" icon and confirm that the "Label Field" is set to "Rt-number."

ArcMap has some nice cartographic goodies that will let us put highway shields on the maps not unlike those you've seen in commercial road maps. Go to Menu View/Toolbars, select labeling tool box. The labling tools like fig 10. will appear in ArcGIS software. Click the second button   to bring up lable manager dialog and input 90 as a new class name and click add (seee table 3-1). A new class with name "90" will appear on the left side. Right click "90" to select it, and select SQL Query to select I-90 from the roads (see table 3-2). In the sql query input window, input as shown in table 3-3, and click ok. Next, click the "Symbol" button , then select the "U.S. Interstate HWY" sheild among many label styles (see table 3-4). Now click Apply to see the the shield to appear on I-90, and click OK to exit the labling tool box.

 Fig 10. Labeling Tool Bar
 Table 3-1. add a new class
Table 3-2. make a query to select I-90.
Table 3-3. the sql query
Table 3-4. select the symbol for I-90

 

Other limited access highways visible in this view are Interstate 93, Mass. State Route 2, and a tiny stub of US 1. Place labels on I-93 and Mass. 2 and ignore US 1. Note that Mass. 2 should not get an Interstate shield since it is a state highway (Admin_type = 3); instead use a round, square, or oval shield. The full set of label tools is shown in Table 3. ArcMap gives you some control over the fonts and colors that appear in the shield symbols. To alter the settings, you may open the Layer Properties window, select the Labels tab and click the Symbol button to open the Symbol Selector window, which gives you the chance to alter the settings. You can experiment with this if you wish, but it is not necessary for this exercise.

Fig. 11. Layer Properties - Lables Tap

Enhanced Layouts

In Lab 1, you created a basic layout and made some simple customizations to it. Now, let's create a more elaborate layout that shows the comparison between the normalized and unnormalized population data. Since you have two data frames, if you create a layout view, Two data frames will be shown in the layout.  Use View > Layout view to create a default layout using the "Portrait" template. Now you can see two frames in the layout. Adjust the size and location of each frames so they are redible. Click the right mouse button while putting the cursor on a data frame to check the properties of the frame.

In this layout we want the two view frames to be the same size and lined up vertically, with "Layer" frame higher on the page than "New Data Frame." Fortunately, ArcMap has tools to resize and align layout elements. Select the Arrow Pointer tool Arrow Pointer tool. Move "View1" near the top of the page and "View2" near the bottom, resizing them, if necessary, so that they both fit. You may also need to move the other layout elements. Now, hold down the Shift key, and click on both frames to select them. Click the right mouse button and choose Align Center from the Align menu.
Align window
Fig. 12. Align Window

Now let's spiff up the frames by adding a border line around their edges. Select one of the view frames and click the right mouse button. Choose properties from the bottom of the drop down menu list. Properties window will come up. Select Frame tab then set the characteristics of border lines as you wish (we recommend keep default settings for this lab.) and click OK. You should see a black line forming a tight box around the view frame.  Repeat this procedure for the other view frame.
Fig. 13. Data Frame Properties - Border Line Setting

Next, you need a legend, a scale bar, and a north arrow to improve the layout view of the map. First, let's insert a legend for the "Layers" frame, Activate the view frame either by clicking it or click the name of data frame(Layers) and then click the right mouse button, and select "activate" from the drop down list. Now select the Insert menu from the menu bar, and select legend. The Legend Wizard window will come up. Using the legend wizard, you can choose which layers will be shown in the legend. For this exercise, select all the names of layers shown in Map Layers space and click ">" button. Selected layers will be shown in the Legend Items space. Click Next. Change the legend title, set font, color and click Next. Now you can choose border line characteristics. Click Next and set spacing between the parts of your legend and click Finish. The legend will show up on the layout view. Follow the same steps, create legend for "New Data Frame". Also, insert scale bars and north arrow for both view frames.
Fig. 14. Select Legend Items Fig. 15. Set legend title

Fig. 16. Set Border Line Fig. 17. Set Spacing between Parts

Finally, round off your map by setting the title of the layout to "Population of Cambridge, MA, 1990" and adding your name, today's date, and appropriate credit to the data source. When you're happy with the way that it looks, save your project, and print out your map.

B. Explanatory Mapping

In this section of the lab we will create a map that makes a specific point. Explanatory maps try to zero in on more specific questions than exploratory maps. Hence, your choice of normalization, classification methods, and the like should be suited to the particular relationship or hypothesis that you postulate. Here we will ask -- and suggest an answer to -- the question, "What is the relationship between land use and home prices?"

Map Land Use and Sales Prices

Open a new empty document and add these layers to it from the M:\data directory: Remember to set the map and Display units for the data frame. Rename this data frame "Land Use and Sales Prices."

Open the table for Landuse85.shp (click once on the layer to activate it and click right mouse button and select Open Attribute Table).. The field called "Landuse" is the only one we're interested in.

Refine the Symbolization

Open the layer properties for Landuse85.shp. Change the "Symbology" to "Unique Value" and select "Landuse" as the "Values Field" then click "Add All Values." Your data is now classified by land use category. Notice that the colors the program automatically chooses are quite random. You can edit them manually, however. Notice in the top right-hand corner of the window are Import button. If you click the Import button, you can either using symbology definitions from other layer files or import ArcView (a previous version of ArcMap) legend file.

Fortunately for you, a legend was previously prepared for this layer using ArcView. Click the Import button, navigate to the usual M:\data directory and select the file landuse.avl (the avl stands for ArcView legend). Click OK. Your layer should pick up the saved symbolization choices. By default, ArcMap orders the categories (values) alphabetically, but we would rather group the residential categories together and have them at the top of the legend. Moving the categories helps us understand them better and will "read" much better on our final map. To move categories, simply click on a symbol and click the up and down arrow buttons in the right side of the window. Locate it where you would like it to be. Move the residential categories to the top:

When you are done click Apply to register your changes.

Open the layer properties for Sales89_shape. Using "Realprice" as your "Classification Field," try out the "Legend Types" "Graduated Color" and "Graduated Symbol." Also try the classification schemes "natural breaks," "quantile," and "equal interval." Remember the question we were interested in? Have these exploratory symbolization exercises shown you the relationship between sale price and land use? It's very difficult to see any type of pattern with so many data points, especially in a place like Cambridge, where high and low income neighborhoods are so close together. We'll now abandon the automatic classification settings and use some of our expertise to determine the classes. We'll look only at very high-priced properties and very low-priced properties. Change the number of classes to "3" in the layer properties - Symbology tab and click Classify button. Change the first number shown in "Break Values" from something like 490196 to 100000 and second number to 1000000. Now you have three categories, less than 100000, more than 1000000, and in between. Click O.K. and Change the color and size of second category (realprice from 100000 to 1000000), so that two extreme values, lower than 100000 and higher than 1000000, stand out. Give the points bright colors that will show up on top of the land use layer. Notice that the high-priced properties are in the lowest density areas. We're on our way to making an explanatory map!

We do not have time to cover all of the symbolization possibilities in ArcMap. The built-in ArcMap help has some useful examples of thematic maps. To see them, select Help Topics from the Help menu. The Help system can take some time to launch. When the ArcMap Help window appears, click once on the "Contents" tab, double-click on the "Creating and Using Maps" book icon, double-click on the "Choosing colors and symbols" book icon, and double-click on "Types of thematic maps."

Creating a Layout

Using the same techniques outlined above in the "Enhanced Layouts" section, create an attractive layout of the "Land Use and Sales Prices" view.

As an added refinement, add a "picture frame" to your layout that includes a CRN and a MIT logo. You will need to use Insert > Picture menu to insert following JPEG files in M:\data:

Make sure to save your map document file before continuing with the next part of this exercise.

III. Exporting Maps from ArcMap

There are a number of reasons you might want to do something with your map other than print it from ArcMap. You may want to create a JPEG or PDF file for use on the web or just get the map into a graphics program to add some finishing touches. Here we consider these two popular formats.

PDF format: This is a format we will use often. Files in this format can be read using a free program called Acrobat Reader. This program comes in a standalone version and as a web browser plugin. The benefit of this format over a JPEG and other bitmap formats is that it's resolution independent. You can zoom in and out of the map and print it at any scale. To create a PDF file from your layout in ArcMap, select File > Export Map. Exoprt window will show up. From the Export window, select the file location, choose "PDF(*.pdf)" option for Save as type, type the File name you want, then click Export.

JPEG format: Follow the same step describe above and choose "JPEG(*.jpg)" as Save as type. Unfortunately, JPEG use "lossy" compression, meaning that a JPEG image is not fully faithful to the original. Artifacts caused by the lossy compression are often visible in JPEG versions of maps. Also, like any bitmap format, JPEGs have a fixed resolution, limiting the ability to zoom in effectively. JPEGs are useful for overview graphics on web pages, while PDFs can be used to supply the full detail.

IV. Lab Assignment

1. Create an exploratory map as described in II A and print it out in Black and White.

2. Create an explanatory map as described in II B and print it out in Color (There is a color printer "Echo" in room 37-312).

Note: All the maps you produce should contain your name, a title, legend, north arrow, scale bar, data source attribution, and, for the Question 2 map, a CRON and MIT logo.

3. Export the two above maps into .pdf format and make it available online.

Since you are at MIT, it's pretty easy publishing your work on the web. As you can see in the explorer, you already have a folder named "www" under I:\ drive. If you save anything in your public "www" folder, then it is automatically up on the web. Now for this class, create a new folder named "11.520" under your I:\www folder. Then save your pdf files in that folder. You don't have to create a good web page for this. Just copy and paste your pdf files into the 11.520 folder you just created. For those who are not MIT students and do not have web spaces, please email pdf files to 11.520staff@mit.edu

Lab 2 is due by the beginning of the class on Sep. 24th, 2008. Hand in your printouts (two maps) to an instructor and send the URL of the two .pdf files to 11.520staff@mit.edu

Created by Thomas H. Grayson.  Includes significant portions originally written by Raj Singh.
Modified for Fall 2000-08 by Thomas H. Grayson, Myounggu Kang, Jeeseong Chung, Jinhua Zhao, Xiongjiu Liao, Mi Diao. Yang Chen. and Yi Zhu.

Last Modified 14 Sept.. 2008 by Joe Ferreira.

Back to the 11.520 Home Page.
Back to the CRON Home Page.

Please send comments about this page to the 11.520 class staff <11.520staff@mit.edu>.