Introduction to GIS and Census Mapping: Part I
by Frank Donnelly, Geospatial Data Librarian, CUNY Baruch College. Version 1.12
Intro to GIS and the ArcMap Interface
Exploring the ArcMap Interface
Let's get started! First, open ArcMap, which is symbolized as a globe with a magnifying glass in front of it. If there isn't an icon on your desktop, look in the Start Menu under Start > All Programs > ArcGIS > ArcMap. Once the software loads, you'll be presented with the empty window in Figure 1.
Your screen may look slightly different from the example, because you may have different toolbars loaded by default. Go to View > Toolbars and check all of the following boxes: Main Menu, Editor, Draw, Layout, Standard, and Tools. For the sake of consistency, drag your toolbars to the same areas of the screen seen in Figure 1. You can drag a toolbar by hovering your mouse over it, hold down the left mouse button, drag it across the screen to a desired location, and release.
An empty window is not much fun, so let's add some actual data. On the top toolbar, click the Add Data button, which has a black plus + sign over a yellow diamond. This will bring you to the Add Data screen in Figure 2:
Now you will have to scan through the directories on your machine to get to your GIS files. Assuming you've stored them on the hard drive, navigate your way to C:\workspace\part1. Once you do, you'll see the geodatabase file in the folder called ResidenceACS_2006.mdb, which is pictured in Figure 2.
! Folder and file names in ArcGIS should be concise, must begin with a letter, and must contain no spaces or punctuation with the exception of underscores _ which are often used instead of spaces.
A geodatabase is simply a container that can hold several GIS layers and attribute tables. It is a nice tool for organizing and storing your data. This particular file, has the extension .mdb, which is the extension for Microsoft Access Databases. Microsoft Access files can serve as Personal Geodatabases in the ArcGIS world. All geodatbases, regardless of type, are represented as grey cylinders in ArcGIS.
We can't add an entire geodatabase to a map; instead, we can add individual features stored inside it. Click on the geodatabase and you will see two features (represented as grey boxes with lines) and two tables (the boxes with blue tops and hashes) as in Figure 3.
In ArcGIS, files inside a geodatabase that represent geographic features are called Feature Classes, and can be displayed visually in ArcMap. The two tables in the database are simple data tables - they have no visual component.
Hold down the shift key and click on the two feature classes metros_2003 and states_2003 and then click OK. Both features will be added to the map as in Figure 4.
Currently, we are looking at two windows in ArcMap. On the left is the Table of Contents which lists all of the features that we currently have added to the map. On the right is the Data View, which gives us a visual, geographic depiction of these files. When you add features to ArcMap, it adds them on top of each other and assigns each one a single, random color.
We can change the drawing order of layers (note - layers is a generic term applied to GIS data files that we will use throughout this tutorial). Hover your mouse over the states layer in the Tables of Contents, hold down the left mouse button, drag the layer below the states layer and release. ArcMap now draws the state layer over top of the metro layer.
Where did the metro areas go? Since the states are being drawn over top of them, they are being covered up by the states layer. This is an important concept to grasp, as you will constantly have to consider what the best drawing order for the layers is. Later in the tutorial, we'll cover how to show both layers at once.
For now, we will not be using the metro layer. You can turn it off by clicking the checkmark button next to the layer's name in the Table of Contents.
Next, let's try some of the zooming features. On the Tools toolbar, you will see a pair of magnifying glasses. Click the first one, which contains the plus symbol. Now, while holding the left mouse button down, click on the NW corner of Washington State, and while continuing to hold the button, drag your mouse over and draw a box around the continental United States. Release the button, and the extent of your view will change to just show the lower 48.
Experiment with the different zoom buttons. Do a single-left click anywhere on the map, and you will zoom in by one click. Use the Zoom Out button, and you can perform the same tasks in reverse. There are also buttons for doing fixed zooms (arrows), one for going to the full extent of your features (the globe), and one for dragging the map around (the hand) which changes what you are focussed on while keeping the same extent. The blue arrow buttons allow you to return to a previous extent (if you zoomed by accident and want to go back) or move forward to the next extent (if you already went back). When you are finished experimenting, focus your view back on the lower 48 states.
As you move your cursor over the map, notice that the units that are changing in the lower-right hand corner. These units are coordinates that correlate areas on the map with real areas on the face of the earth. Remember that GIS features are not simple drawings - they are georeferenced data layers. The units you see in the corner will vary based on the map projection and coordinate system your layers are in.
At this point, let's save our map. Go up to File and select Save As. Save your map in the same directory as your geodatabase. Map files in ArcGIS are identified with the extension .mxd.
What's actually being saved in the mxd file? The following items: A list of all the layers and tables in the Table of Contents with their order and status, the current view in the Data View window, the symbolization and classification of all of your layers, table, joins, selections, and all map elements in the Layout View (which we'll cover later). What isn't saved in the mxd? Your actual GIS layers and tables. They are not stored in the mxd file. Rather, the mxd file links to these files, which are stored outside either in folders or in geodatabases. You can think of the ArcGIS interface as a window for viewing your data, and the mxd file saves all of the window dressing. When you are changing how data is displayed in your map, you are not actually changing the data files themselves. You are simply changing how you are viewing them within your mxd file.
!Since your mxd files are linked to the data files, the mxd looks at the physical location or path of these files (i.e. C:\Workspace\Census_Part1\censusmap.mxd) each time you open your map so that it can display them. If you move your files around or change their names, this link will get broken. It's important to think about how to organize your data before you begin, because you will run into problems later on if you don't. We'll look at how to restore broken links a little later.
!In particular, this problem occurs if you are saving projects on a flash drive. When you plug a flash drive into a machine running Windows, Windows assigns it a drive letter. However, this is not consistent from machine to machine - one machine may assign it G: and another may assign it F:. If ArcGIS is looking for your files at G:\Workspace\Census_Part1\censusmap.mxd and you move to another machine that looks at your flash drive as F:, ArcGIS will not be able to find your files, and all your links will be broken. You can avoid this problem by going to File > Properties > Data Source Options and specify that ArcGIS stores data path with relative names instead of fixed names. That way, ArcGIS will only look in the folders and subfolders where your mxd files are stored (i.e. \Census_Part1) and will not look above to higher folders and the drive letter. There is a caveat to this method too - you have to store your data in the same directory or in subdirectories below your mxd files - you can't store them in separate unrelated folders, otherwise ArcGIS won't be able to find your files!
Identifying and Selecting Features
Let's identify some of the features on the map. Click on the Identify tool in the Toolbar (the blue circle with the letter i inside it). Next, click on New York State in the data view. Once you do, you'll see a window with some information about New York.
Where is this data coming from? All vector-based features in ArcGIS (features that are made up of points, lines, or polygons/shapes) have attributes associated with them that are stored in an attribute table. The identify tool enables us to quickly see that information for a particular feature. In this case, we can see ID codes, the name of the state, length of its boundaries, and area. The units for length and area are in the units used by the map projection. In this case, it is meters for length and sq meters for area.
We can view the entire attribute table for a feature class. Do a single right click on the states layer in the Table of Contents to bring up a menu for that layer.
Select the Open Attribute Table option to open the table. Now you should see a table with one record for each state, and with columns showing several different attributes. Scroll down until you see New York in the table. Click on the small, grey box in the New York State row, just beside the [OBJECTID] field. This will hi-light the record in the table, and, will hi-light New York in the Data View. Reduce the size of the table window so that you can see both the table and New York in the data view.
You can select multiple states by holding down the control (ctrl) key. Hold down the ctrl key and select New Jersey. You should now have both NY and NJ selected in the table, and you can see them both selected in the data view. To clear selected features, go to Selection and choose Clear Selected Features.
This operation also works in reverse - you can select features from the data view and they will be hi-lighted in the table. Move your table aside, and in the data view in the Tools toolbar, locate the Select Features button. It has a white arrow beside a blue and white box. Then point to the State of Pennsylvania and click it. Now, look at your table. Pennsylvania should be hi-lighted.
Essentially, we've just demonstrated that feature classes consist of two components that we can see - visual geographic features and tabular data features - and we can work with either the features or the data. This is one of the great strengths of GIS will allows us to view and manipulate our data in ways that other types of software cannot.
Each of our feature classes have other properties that we can view and manipulate. Let's start to explore some of these. Single right-click on the states layer, and select Properties from the menu. You should see a window that looks like figure 12.
The Properties window has several different tabs that relate to aspects of our data. By default, your window should open to the General Tab as shown in the figure above. There's nothing particularly exciting about the General Tab. But the Source Tab on the other hand is quite relevant. Click on the Source Tab and examine its contents.
The Source Tab provides us with two important pieces of information. First, it tells us where this particular feature class is stored. Remember, the states layer is NOT stored within this particular map that we are working with. It is an independent file sitting in a geodatabase, and our map is simply pointing to the database and is showing us the file. If the link between our map and the geodatabase was ever broken (because we moved either the map or the geodatabase to a different folder or renamed a file), we could come to this window and restore the link.
Second, the Source Tab tells us what projection our feature class is in and what coordinate system it uses. In order for all of our layers to display and work correctly in a map, they must all share the same projection and coordinate system. Our states feature class uses the USA Contiguous Albers Equal Area Conic projection, which is commonly used for making maps of the Lower 48. It uses the GCS North American 1983 as a coordinate system. GCS stands for General Coordinate System, which is latitude and longitude based.
Symbolizing and Labeling Data
Next, click on the Symbology Tab. You will use this tab quite often - it is where you change how your features and data will be displayed in the data view.
By default, ArcGIS displays all of the features in a layer using a single symbol, which is rather boring. However, we can display symbols in a variety of ways, based on the attributes stored in the attribute table for the layer. In the list on the left-hand side of the symbology tab, click Categories.
In the Categories section, we will be able to display each state as a different color based on some category. The Values dropdown box includes a list of all of the fields in the attribute table for states. Pull down this box, select [NAME], and hit the Add All Values button. This will give each state a different color based on it's name.
So in this case, each category will have one value - one state for each category. But you can imagine if we had an attribute field that specified what region a state was in - Northeast, South, Midwest, etc, then we would have multiple states in each category, and every state within a category would receive the same color. If we decided that we did not want to map certain categories (in our example, certain states), we can remove values from the list one by one, or we could have chosen to add specific values rather than all of them. ArcGIS also gives us an All Other Values category to symbolize values that don't have a category or values that we have chosen not to display. You can uncheck this box.
Use the color ramp drop down menu to change the color choice. Colors are assigned to category values at random. Once your screen resembles the window in Figure 16 above, hit the OK button, and you will be presented with a colorful map.
If you are not happy with the arrangement of colors, go back to the Symbology tab and reselect the color ramp. If you select the same color ramp as before, it will randomly assign the colors again, and you may end up with a better mix. If you want to be particular, you can change the color for an individual feature. In the Table of Contents, doubleclick on the color box beside Florida. This will bring you to the Symbol Selector window depicted in Figure 18.
Each symbol has a fill and an outline. Use the fill color dropdown box to choose a new color for Florida. We'll leave the outline color and width alone for now. Click OK, and Florida will change colors before your eyes.
Next, let's put some labels on the states (in case you don't have them all memorized). Single-right click on the states layer in the Table of Contents (make sure you do this on the layer name and not on an individual state) and go to the Properties. Click on the Labels Tab, and you'll get the window in Figure 19.
ArcGIS can assign labels to features based on attributes stored in the attribute table. In the label field dropdown box, select [NAME]. Then, check the checkmark box in the upper left-hand corner that says Label Features In This Layer. Then click OK, and voila - the states are labeled with their names.
As you can see in Figure 20, our result isn't perfect. ArcGIS tries to place the labels in optimum places, but doesn't always succeed if you have many features and are squeezing too many labels into a small space. In this case, the font is a little too large, and Delaware and Rhode Island were omitted because of conflicting labels. Michigan on the other hand was given two labels, one for the upper peninsula and another for the lower half. Fortunately, we can sort this out.
Go back to the Labels Tab in the Properties Window. Change the font size to something smaller. Then click on the Placement Properties button, and on the Duplicate Labels Section (shown in Figure 21), select Remove Duplicate Labels. Hit OK on both screens, and our map should be in better shape.
Zoom in to the Mid-Atlantic section of the country, and watch what happens to the labels. ArcMap redraws the labels every time you change the extent (or scale) of your view, so that they are optimally placed each time.
ArcGIS has several features and options for working with labels to address issues like optimal placement, conflict detection, and label size versus map scale - too many features for us to cover in this brief tutorial. We'll revisit label placement later on.
Adding and Joining Data Tables
Thus far, we have created a simple, working reference map that shows the boundaries of the states. Reference maps, as the name implies, are used for simple reference - they show us where things are. Now, let's move on to making a thematic map. As the name suggests, a thematic map displays themes about places. Instead of just telling us where things are, they describe what's going on in a particular place, and if there is any pattern to what's going on.
Click the add data button (black plus sign on yellow field) and go back into our geodatabase. Look for the table called states_residence, select it, and add it to our map.
After you add it, nothing special seems to happen. The table we added is just a plain old data table that has no visual geographic component. ArcMap opens a new tab at the bottom of the Table of Contents window called Source, where we can see that the table is added to our map. It does not appear in the display tab though (where all of our feature classes appear), because it has no visual, geographic element. It's just a bunch of rows and columns.
Right click on the table to see the Properties menu, and open it by selecting Open Attribute Table. Figure 24 illustrates what it should look like.
This table contains data from the 2006 American Community Survey (ACS for short) for residence. Households were asked if they lived in the same house a year ago. The first column shows the total estimated population of each state. Then there are 5 categories - Yes, I did live in the same house, No, I moved from a different place within the same county, No, I moved from a different county within the same state, No, I moved from a different state, No, I moved from a different country. There are three data columns for each of these categories: one for the total population over 1 years of age, one for men, and one for women. For every single data column, there is also a margin of error column. Since these are estimates from a survey (and not exact counts), the Census Bureau provides us with a margin of error. For example, we would read the total population estimate in field [B07003_1_EST] for Alabama as "We are 90% confident that the population for Alabama in 2006 is 4,542,703, plus or minus 3,318 people."
A code is used to name each of the columns. You need to use a look-up table in order to determine what each code relates to. There is an Excel file in the Part I folder called ACS_2006_Res_Codes.xls that relates each code to a name. This seems cumbersome, but it would be more cumbersome if we had the full name of all the variables in every single field. Imagine squeezing the name of the variable for field [B07003_13_EST] into the column heading: Universe: POPULATION 1 YEAR AND OVER IN THE UNITED STATES: Moved from different state (Estimate).
! Names for column / field headings follow the same rules as names for folders and files. The names must be short with no spaces (use underscores instead) and no punctuation. Names cannot begin with numbers - they must begin with letters. If you import a table that violates these rules, you will encounter null values and other errors.
Even though this is just a simple table, this table does contain geographic data. Each record represents a place - in this case, a state. And we have a column that uniquely identifies each state. The [GEO_ID2] column contains the two-digit FIPS code for each state. FIPS code are used by the US government to uniquely identify each place in the US.
Now, open the attribute table for the states feature class, and examine it next to the ACS table.
In the attribute table for the states, there is a field called [STATE] which also contains the FIPS code for each state. This is good news! ArcGIS functions like a relational database. In a relational database, if you have two tables that have different data sets for the same items, and each of these tables share a common identifier like a code number, you can join these two tables together so you can work with the data from both sources at once. In this case, we have two tables with data about states. The attribute table for the GIS layer has geographic data about each state. The ACS table has residence data about each state. Each table uses the same code, a FIPS code, which uniquely identifies each state (the columns may have different names, but they both contain FIPS codes). So, we can use ArcGIS to join these two tables based on the FIPS codes, which will allow us to map the data in the ACS table.
Let's join the tables. Close them, and then right click on the states layer in the Table of Contents and select Joins and Relates > Joins from the menu, as shown in Figure 26.
In the Join Data Window, Join Attributes From a Table should be the option in the first drop down box. In option 1, choose the [STATE] field as the the field in the states layer that you want to use for the join (since it contains the FIPS code). In option 2, choose the states_residence table as the table you want to join to the states layer. In option 3, choose the [GEO_ID2] field as the field in the states_residence table that you want to use for the join (since it contains the FIPS code). ArcGIS will attempt to match the two tables and join them based on these two fields. Click OK
! Each table also has a field with the name of the state - couldn't we use this field to do the join? Generally, it's a bad idea to use names of places to join tables. Names can be spelled differently, abbreviated, abbreviated differently, and spelled with caps and lowercase, which means they may not be identical in both tables. In other cases, we may have problems with toponyms - the same name for two different places (Manhattan, NY or Manhattan Kansas?) It's better to use something consistent like a code. In this case, the FIPS code is the best choice. If we had the two letter postal codes (NJ, NM, NY, etc.) those would have been acceptable as well.
Once you do a join, nothing visible happens. To see if it worked, open the attribute table for the states layer. All of the columns from the ACS table should be tacked on the end of the attribute table for the states. For each heading, ArcGIS puts the name of the file or table before each of the field names, so you know that the data in the joined table is coming from two different sources.
! Are you seeing null values for all of the ACS columns in the attribute table? Remove the joins by going back to Joins and Relates and Select Remove All Joins. Then, try to do the join again, and verify that you have used the correct fields and tables in the Join Window. If this still does not work, you will have to check the properties for each of the join fields to make sure that they are both saved as text data types. You can only join fields that have the same data type. If one is stored as text, and the other is stored as a number, you'll have to convert one of the fields. Another possibility: check the names of your columns. If they violate the naming rules because they have strange punctuation, spaces, or begin with numbers instead of letters, you'll have to change the names and try again.
Now that the tables are joined, let's map some data! Open the properties for the state layer and go to the Symbology tab. This time, select Quantities instead of Categories from the menu on the left (see Figure 28). In the Values drop down menu, you will see a list of all of the fields in the attribute table, including the fields that have been joined from the ACS table. Select field [B07003_13_EST] from the drop down. Click OK.
Ta-da! We now have a thematic map, that shows us the number of residents of each state that moved within the last year from another state. While this is a dazzling start, we could do better. States with a higher population will naturally have a higher number of residents regardless of the category, so this map doesn't tell us much.
Go back to the Symbology tab. In the drop down menu for Normalization, select the field for Total Estimated Population, [B07003_1_EST]. The field in the Value dropdown will be divided by the field in the Normalization dropdown, which will give us a percentage rather than a total value.
Let's fix the labels so they show us percentages without adding tons of digits. Single-left click on the Labels title and go to Format Labels. In the Number Format window (Figure 30) go to the percentage category and select the radio button that says that the values are fractions, turn them into percentages. Then hit the Numeric Options button and change the value for decimal places to one. Hit OK to return back to the symbology tab.
Once you are back to the Symbology tab, under the classifications dropdown, change the value from 5 to 4. This will give us four data categories rather than five. Once your screen looks like Figure 31, click ok.
Now we have a map that is a little more meaningful, that shows us the percentage of each state's residents who have moved to that state from another part of the country within the last year. We can see that many people have moved to the mountain states of the west and to the southeast coast from other states, while states in the northeast, upper midwest, and California have seen fewer migrants from other states.
How were the four categories created? We chose the number of categories ourselves, by changing the default from 5 to 4. If you use too few or too many categories, it becomes difficult to visually detect patterns. It's best to stick with 3 to 6 categories, depending on what you're mapping. By default, ArcGIS uses a classification method called Natural Breaks. This method statistically analyzes your data, looks for gaps between the values, and creates categories based on where these gaps are (or if you prefer, where data is clustered together). This is a sound method to use, but there are other methods to choose from. Let's look at some other options.
Open the Properties menu for the states layer and return to the symbology tab. Click on the Classify button.
The Classification window gives you some stats about your data, and shows you where values fall within each class. If necessary, you can use this window to manually change the categories and to exclude outliers. You can also select different classification schemes. Under the Method dropdown box, choose the Equal Interval option and watch how this changes your distribution. Click OK here, and OK again on the symbology tab, and view your map.
We still see a similar pattern, but the differences are a little less pronounced. What happened here? The Equal Interval scheme takes the range of your data values and divides them into equal parts based on the number of classes you chose. The lowest value is 1.4% and the highest is 8.5% (Washington DC is the one high value - we can't really see it well on our map). Each of our categories contains approx 1.8 percentage points (3.2 minus 1.4 is 1.8, 5 minus 3.2 is 1.8, etc.) Let's take a look at one more classification scheme.
Return to the Symbology tab via the Properties window. Hit the Classification button and change the method to Quantiles. Hit OK, OK again, and view your map.
We still have a similar pattern, but here the differences are more pronounced. The Quantiles scheme takes the number of data points that you have and divides them into equal parts based on the number of classes you chose. There are 51 data points. Divide them by 4, and you get approximately 13 points per category, sorted from lowest to highest.
Each of these methods takes the same data and the same geography and gives you a different outcome. Which is the best method to use? There is no simple answer. It really depends on the distribution of your data, and on the message you're trying to send with your map.
Working With Multiple Layers
One of the features that makes GIS unique is its ability to take different features and layer them over top of each other. So far, we've been working with one layer. Let's add another layer into the mix. Turn on the metromicro_2003 layer. Then, in the Table of Contents, drag this layer over top of the states layer, so that we can see it (it's currently being covered up by the states layer).
The result here looks rather busy. We're going to focus on working with this new layer, so let's turn off some of the features of our state layer. Go back to the Symbology tab for the states layer, and choose Single Feature. Click on the solid feature color in the window to edit the properties for the color. Under the Fill Drop down, select No Fill Color. Then change the outline width to 1.0 and change the outline color to a light grey. Then click OK.
Next, right click on the states layer in the table of contents, and under the menu select the option for turning the labels off. We've now uncluttered our map, and have changed the state layer's purpose. It's now a simple reference layer, rather than a thematic one.
Our new layer, metromicro_2003, contains Core Based Statistical Areas (CBSAs), which are urban areas created by the Census Bureau. CBSA's consist of two types of areas: Metropolitan Areas and Micropolitan Areas. According to the Census Bureau :
"A metro area contains a core urban area of 50,000 or more population, and a micro area contains an urban core of at least 10,000 (but less than 50,000) population. Each metro or micro area consists of one or more counties and includes the counties containing the core urban area, as well as any adjacent counties that have a high degree of social and economic integration (as measured by commuting to work) with the urban core." - US Census Bureau Metropolitan and Micropolitan Statistical Areas http://www.census.gov/population/www/estimates/metroarea.html (3/18/2008)
Let's map some residence data based on metro areas. Hit the add data button and navigate back to our geodatabase. Add the metros_residence table to our map. This table contains the same kind of data that our states_residence table had, except it pertains to metropolitan areas. Open the metros_residence table and the attributes of the metromicro_2003 layer, and let's compare them side by side.
If we look at the bottom of each table window, we can see how many records we have in each table. In our metros_residence table, we have 367 records, and in the attribute table for our metromicro_2003 feature class we have 935. So, we have a problem here, because our data table and our feature class do not contain the same number of records - the feature class has a lot more. Can you determine what the difference is?
The difference is, the feature class has geography for metropolitan and micropolitan areas, whereas the data table only has records for metropolitan areas. Can we join these tables together? We could. Since there is no micropolitan data in the data table, all of the data fields for micropolitan areas in the feature class would be populated with null values (blanks) following a join. A better solution would be to use a feature class that has just metropolitan areas. Do we need to go out and find one? No! We can just modify what we have. The feature class has a field called [TYPE], which indicates whether each feature is a metro or micro area. We can use this to filter out the micros and keep the metros.
In ArcMap, go up to the Selection Menu listed along the top of the screen and choose the Select By Attributes Option. This will open the window shown in Figure 39.
Think back to when we were selecting features. Remember that we could select features by clicking on them in the attribute table or on the map. The Select By Attribute window will let use select features based on values stored in their attribute table.
In the Select by Attributes window, select the metro_micro2003 layer in the Layer dropdown box. Leave the method as Create New Selection. Under that you will see a list of attribute fields in this layer. Double click on the [TYPE] field. This will make the [TYPE] field appear in bottom window, and it will cause the different values that appear in the [TYPE] field to appear in the middle window. Hit the equals sign = on the calculator, making it appear in the bottom window to the right of [TYPE]. Then click on the Metro value from the middle of the list, which causes it to appear in the window below. Read the statement that appears above the bottom window and read the statement that appears in the bottom box. All together, it reads: SELECT* FROM metromicro_2003 WHERE [TYPE] = 'METRO'. Then hit OK to run the query.
The statement you just created is called a SQL statement. SQL is the Standard Query Langauge that is used in all databases. You told ArcGIS to select all of the features in the metro_micro layer that are categorized as metropolitan areas, and you should get the result you see below:
Now that all the metro areas are selected, we can create a new, exclusive layer out of them. Go to the Properties menu of the metromicro_2003 layer and select Data > Export Data as seen in figure 41.
This will bring you to the export data window. You will export the selected features. Click on the little folder to specify where you'll want the new file to go. In this case, put it inside our geodatabase file, and give the new feature class the name metros_2003.
Click OK, and ArcGIS will export the features to our geodatabase. When this is complete, respond yes to the prompt and add the new feature to the map. Since we don't need the old metromicro_2003 layer, we can remove it from our map. Go to the Properties menu for that layer and select remove. This doesn't delete the layer (it's still stored safely in the geodatabase), it just removes it from our map.
Now that we have a metro feature class to go with our metro data table, go ahead and join the two. The FIPS code for the feature class is in the [GEOID_2] field, and the FIPS code in the data table is stored in a field called [CBSA]. Get the Properties menu for the metro feature class, Joins and Relates, Join. Follow the same steps we used previously for doing the join, and check the attribute table when you are finished.
Once the data has been joined, go to the Symbology tab for the metros. Select the [B07003_13_EST] field as the Value and the [B07003_1_EST] field as the Normalization Field. Choose 4 classification categories and alter the labels like we did before, to show percentages with one decimal place. Once you've finished, your symbology menu should look like Figure 44:
Once you are ready, click OK to see your map.
This is a pretty decent map, which shows you the percentage of each metro area's residents who moved there from another state within the last year. The areas are large enough that you can clearly see the shading within each. So far, we have been using shaded areas to map our data - the fancy name for this is a choropleth map. However, we do have other options at our disposal besides shaded areas for depicting our data.
Symbolizing Data - Part II
Go to the Symbology tab for the metro layers. Under the Quantities list choose Graduated Symbols. In the symbol size boxes, change the size values to 5 for the minimum and 20 for the maximum. The template button on the right lets you choose a shape - we'll keep circles for now. Click on the area button below it and select options for no fill and no line. That way, we will hide the actual boundaries of the metro areas and will be left with just circles for each one. Click OK. Change the circle color to something you like. Your window should look like Figure 46.
Once things look good, click OK to see your new map.
When you are working with geographic areas that are contiguous and cover the entire extent of your map (like the states layer), it is common to use shaded areas. When mapping non-contiguous areas (like the metro areas) or geographic areas that are saved as point layers (rather than polygons or shapes), it is more common to use graduated symbols. In this particular case with the metro areas, you could go either way. It is difficult to see all of the metros at this national scale, as many of the circles overlap. If we were interested in looking at just a subset of these metros, say, just the largest ones, then graduated symbols would be a little more feasible.
Go back to the symbology tab for metros layer, and hit the Classification button. From the Classification window, hit the Exclusion button to get the window in Figure 48.
This window looks just the the Select by Location window that we previously used. However, the result here will be different. Rather than just selecting features, this window will allow us to exclude features from being presented in our classification scheme and on our map. Click on the [B07003_1_EST] field (total population), select the less than symbol, and type in 500000. Hit OK on each screen to get back to our map.
Now, we just have metros with a population of over 500,000 people visible, and our classification scheme has changed to account for the fact that they have been excluded. They are still there, stored in the table; we're just choosing not to view them. We also have a map that's a little less cluttered, and which effectively communicates the pattern of new migrants in large metropolitan areas.
So far, we have been working in the Data View in ArcMap. If we want to create an actual finished product, a map in digital or paper form, we need to use a different set of tools in a different view - the Layout View. Let's move on to Part II of this tutorial, Making Maps in the Layout View.
Begin Part II.