Section3 Exercise4 ThematicMapping
Section3 Exercise4 ThematicMapping
MOOC
Exercise
Thematic Mapping
Section 3 Exercise 4
January 20, 2021
Cartography. MOOC
Thematic Mapping
Instructions
Use this guide and ArcGIS Pro to reproduce the results of the exercise on your own.
Note: The version of ArcGIS Pro that you are using for this course may produce slightly
different results from the screen shots that you see in the course materials.
Time to complete
Approximately 50-60 minutes
Software requirements
ArcGIS Pro 2.7
ArcGIS Pro Standard license (or higher)
Note: The MOOC provides a separate ArcGIS account (user name and password) that you
will need to use to license ArcGIS Pro and access other software applications used
throughout the MOOC exercises. This account (user name ending with _cart) provides the
appropriate ArcGIS Online role, ArcGIS Pro license, ArcGIS Pro extensions, and credits.
We strongly recommend that you use the provided course ArcGIS account to ensure that
you have the appropriate licensing to complete the exercises. Exercises may require
credits. Using the provided course ArcGIS account ensures that you do not consume your
organization's credits. Esri is not responsible for any credits consumed if you use a different
account. Moreover, Esri will not provide technical support to students who use a different
account.
Introduction
Special purpose maps, as opposed to general reference maps, are termed "thematic maps."
These maps tend to display a single theme of information by indicating the spatial distribution
of a single attribute or the relationship among many attributes. Thematic maps illustrate the
structural characteristics of the geographical distribution of a dataset.
Thematic maps vary considerably, but examples include maps depicting precipitation,
population distribution, hurricane incidence, and pollution. Thematic maps tend to be small
scale (showing a large area, which is expressed as a small representative fraction) because the
geographical distribution of a theme is often most appropriately shown and accurately
depicted over a large area.
Thematic mapping requires the transfer of empirical measurements into graphical signs that
can be mapped to illustrate the spatial variation in the difference between measured values.
Essentially, aspatial, tabular data is converted into a spatial form that, by definition, adds
further to the information by giving the map reader the distance, direction, shape, and
location of the theme. In a thematic map, the transformation of data from tabular to map form
adds spatial understanding.
In this exercise, you will develop a range of alternative thematic maps that tell a story of the
2015 general election in the United Kingdom. To see various types of thematic maps, you
should look at the ArcGIS Online thematic mapping gallery (https://bit.ly/2GQhgC6), which
shows a range of different ways of mapping the 2016 U.S. presidential election and includes
discussions of the map type and use.
For your reference, the project package includes a map view containing the final maps for this
exercise. Your final maps may not match exactly. You are encouraged to experiment with the
options available in ArcGIS Pro.
a Start ArcGIS Pro and, if necessary, sign in using your provided course ArcGIS credentials
(user name ending with _cart).
b From the main ArcGIS Pro start page, click Open Another Project.
c Browse to the location where you saved the exercise data file and open the
ThematicMapping.ppkx project package.
The project opens with a map showing electoral constituencies for the 2015 UK Election.
d Save your project with a name such as ThematicMapping_<your first and last
name>.aprx in the folder on your computer where you are saving your work.
This dataset is the same one that you used in the Section 2 exercise Exploring Data for
Classification, so you should be familiar with its general contents. From that exercise, you will
also use data classification concepts to process some of this data for mapping.
A single layer in the Contents pane shows the electoral constituencies of Great Britain and
Northern Ireland. Additionally, a version of the data represents the election results as points,
each in the center of the constituency. Whether you use the area or point version of the data
depends on the map type, which is addressed later in this exercise.
a Right-click the Constituencies layer and choose Attribute Table to review the fields and
their contents to refamiliarize yourself with the data.
The layer attribute table contains a number of fields, which are described in the following
table:
b When you are finished exploring the data, close the attribute table.
First, you will make a simple thematic map showing the winning political party in each
constituency.
b Right-click the Constituencies layer and choose Symbology to open the Symbology pane.
c In the Symbology pane, from the Primary Symbology drop-down list, choose Unique
Values.
Note: The colors are assigned automatically, so the colors on your map may be different than
the ones in the preceding graphic.
This data is a text field that identifies the abbreviation of the political party that won that
constituency. The result is a basic, descriptive map of the result of the election.
Default colors and labels rarely make an elegant map (as you can hopefully see). First, you will
update the labels to make them more meaningful.
Note: You can dock the Symbology pane next to your map by right-clicking the pane title and
choosing Dock. To auto hide the Symbology pane, click the Auto-Hide button at the top
of the pane.
e If necessary, at the bottom of the Symbology pane, scroll to the right to view the Label
column in the grid.
Note: Alternatively, you can make the Symbology pane larger by dragging its border.
f Under the Label column, double-click each label and change it to the appropriate political
party label, as shown in the following table.
Value Label
CON Conservative
DUP Democratic Unionist Party
GRN Green
IND Independent
LAB Labour
LD Liberal Democrats
Other Other
PC Plaid Cymru
SDLP Social Democratic and Labour Party
SF Sinn Fein
SNP Scottish National Party
UKIP UK Independence Party
UUP Ulster Unionist Party
The party names are now updated in the Contents pane. Remember, it is vital that your map
reader can interpret the map, and the way that data is stored is often not in an immediately
recognizable form.
It is useful to make use of sensible symbology to reinforce the map's message. For instance,
for this dataset, the default uses randomly selected colors, yet political parties have their own
color schemes. This project comes preconfigured with a style file of colors that are ready to
use, but you can always use a color picker to change them to what you prefer.
Next, you will change the colors for each of the political parties to match their party color
from the Political Colours style file.
g In the Contents pane (not the Symbology pane), right-click the color chip for Conservative
to open the color picker.
h In the color picker, in the Political Colours group, click the blue Conservative color.
Hint: Point to a color chip to see its name.
i Repeat this process to assign party colors based on the party name.
When you are finished, the symbols for the parties should look like the following graphic:
j In the Symbology pane, click More and uncheck the box next to Show All Other Values.
You will also change the color of the outline of each symbol so it better matches the
basemap. It is always worth remembering that changing the outline of your area symbols is
often as important as the fill color itself. Sometimes, you may even make the outline
transparent or with a width of 0 pt to make it disappear, which can give a very nice effect and
avoid a lot of lines getting in the way.
k In the Symbology pane, click More and choose Format All Symbols.
l In the Format Polygon Symbols pane, from the Properties tab, change the outline color to
Gray 80% to better match the dark gray basemap.
m Click Apply.
o Zoom and pan the map to explore how this thematic map effectively shows the winning
political party in each area using a unique value (winning party) symbolized by the political
party color.
Much better than the default, right?
You will now make a simple graduated colors (choropleth) map to show the share of votes that
the winning member of Parliament (MP) received. This data is in the First_Share field in the
Constituencies layer attribute table.
a In the Contents pane, right-click the Constituencies layer and choose Copy.
c Rename the layer that you just created to Winning share of vote - choropleth.
e Open the Symbology pane for the Winning Share Of Vote - Choropleth layer, and then
update the primary symbology to Graduated Colors.
Graduated colors are used to symbolize the layer by quantities. In this case, the quantity is the
percentage or share of votes that the winner received.
Do you recall from the Section 2 exercise Exploring Data for Classification that you cannot use
a choropleth map to map totals? You are fine here, however, because you are mapping data
that has already been normalized by converting the raw totals to percentages.
Your final choropleth map should resemble the following graphic.
You now have a default choropleth map showing the share of the vote received by the
winning MP classified into five classes using the natural breaks method. Lower values are
shown in lighter colors (a lighter hue), and higher values are shown in a darker color (a darker
hue).
Challenge: Experiment with data classification and colors
In this optional challenge, create a map that shows enough variation to effectively represent
the results. With too few classes, the map will be too general. With too many classes, the map
will be too complicated. A good guide is four, five, or six classes.
Hint: Refer to the Section 2 exercise Exploring Data for Classification for a reminder of the
data classification options. You can also refer to ArcGIS Pro Help for more information about
data classification options (https://bit.ly/2qnFLzl).
You can use your default choropleth map (make a copy of the Winning Share Of Vote -
Choropleth layer and name the new layer something like Choropleth - challenge) to
experiment with the following:
• Different classification types (available in the Symbology pane from the Method drop-
down list)
• Number of classes (available in the Symbology pane from the Classes drop-down list)
• Color schemes (available in the Symbology pane from the Color Scheme drop-down
list)
For a choropleth map, the best way to represent data that varies from high to low is to use a
color scheme that visually reflects that. A single-hue color scheme works well. A spectral color
scheme does not work well because there is no intuitive way to determine whether one hue is
"higher" or "lower" than the other. This issue is referred to as the classic "rainbow color
problem."
• In the Symbology pane, in the Label section, change the way that the labels appear to
remove the trailing zeros and clean up the legend.
You could change the style so that you have an upper and lower label such as 0-40%, 41-50%,
and so on. Avoid overlapping classes (where you use the same label at the higher end of one
class and the lower end of the next class).
a In the Contents pane, verify that the Constituencies layer is turned off and turn off the
Winning Share Of Vote - Choropleth layer.
b If you completed the challenge at the end of the previous step, also turn off that layer.
c Right-click the UK Election 2015 map item and choose New Group Layer.
f In the Value-By-Alpha group layer, paste two copies of the Constituencies layer.
i Right-click the Turnout layer and choose Symbology to open the Symbology pane.
j For Primary Symbology, choose Graduated Colors and update the Field to Turnout.
k For Color Scheme, choose the White To Black gradient color scheme.
Note: If you do not see the White To Black gradient color scheme, you may need to click the
Color Scheme down arrow and check the box for Show All.
l In the Symbology pane, next to Color Scheme, click the down arrow and choose Format
Color Scheme.
m In the Color Scheme Editor, change the black color chip to be 100% transparent.
Hint: Click Color and choose No Color.
o If necessary, in the Contents pane, in the Value-By-Alpha group layer, turn on the
Constituencies layer.
For a well-balanced result, you can use the various controls in the Symbology pane to adjust
how the white-transparent color ramp stretches across the data values.
The map brings into focus constituencies that had higher turnouts. The top layer in the group
effectively mutes areas where there were lower turnouts. This useful technique can show how
two related pieces of data interact with one another.
a In the Contents pane, turn off all layers except the Constituency Points layer and the Dark
Gray Canvas basemap layer.
c Change the symbology type to Graduated Symbols and use First_Votes as the field to
symbolize.
The default classification is Natural Breaks (Jenks) using five classes. Every data value falls into
one of these symbol classes.
d Using the Symbology pane, experiment with changing the color fill, outline, and sizes of
the symbols.
One of the challenges with any symbol-based map is congestion in parts of the map.
Balancing the symbol size with the appearance of the overall map is a good goal to keep in
mind. Transparent symbols with a solid outline can sometimes work, so include those in your
experimentation.
Also, notice how as you zoom in and out of the map, the symbols remain the same size at
which you designed them; as you zoom in to a larger map scale, your overlapping symbols
tend to become separated from one another. This useful feature of a multiscale mapping
environment counteracts the problems of seeing overcrowded symbols at smaller scales.
If you prefer that your symbols stay at their relative size as you zoom in and out, you can
achieve this result by setting a reference scale (https://bit.ly/34RkIsV).
e In the Contents pane, right-click the UK Election 2015 map item and choose Set
Reference Scale.
f Right-click the Constituency Points (the layer symbolized with graduated symbols) layer
and choose Properties.
g In the Layer Properties dialog box, from the Display tab, ensure that the Scale Symbols
When A Reference Scale Is Set box is checked.
h Click OK, and then zoom and pan the map to see the result of changing the reference
scale settings.
Effectively, this setting freezes the symbol sizes at the scale at which the reference scale was
set. As you zoom in and out, they become larger and smaller while maintaining their relative
size to the map and each other symbol.
i Clear the reference scale by right-clicking the UK Election 2015 map item and choosing
Clear Reference Scale.
Clearing the reference scale will take you back to the default for the next step in this exercise.
h Use the First_Votes field as the field to symbolize again or experiment with other fields.
In this map, the data value is scaled to the symbol size, so each symbol on the map is a
proportional representation of the data. Every value of data is represented uniquely, so this
map type shows much more variation in symbol sizes than a graduated-symbol map.
i Experiment with the scaling settings and symbol design to achieve a nice visual balance.
You are trying to achieve a balance where there is enough variation in size across the map to
show highs and lows in the data but without symbols becoming either too large and
dominating or too small and difficult to see. It is not a perfect science, and experimentation is
important to go back and forth until you feel like you have captured the optimum settings for
the dataset at hand. Importantly, ask yourself what you want the map to tell the reader. The
answer to this question should help you consider whether your scaling settings and symbol
design have met that goal.
One final aspect to scaling settings to consider is whether you need all your symbols to be
scaled uniquely. In the histogram view of the Symbology pane for this layer, you can move the
sliders to change the way scaling is applied.
j Move the top slider (lowest-level data value) down, as shown in the following graphic.
Updating the slider will lead to all data values below the threshold that you set being shown
at the same size. It can be useful to highlight variation at one end of the data array, or
conversely, if you have outliers (https://bit.ly/2VIciQC) or many data values of an extremely
small or large size, you can cluster them together and show them (and them only) with a
symbol of the same size.
k Zoom in to view of the final proportional symbols thematic map based on the First_Votes
field values.
Now you will modify the symbol color to make each of your symbols in a single layer
represent more than one data item using attribute-driven symbology. Taking the Proportional
Symbols map that you just made, you will make it multivariate.
b Paste the copy in the UK Election 2015 map item, and then rename it Multivariate
Proportional Symbol.
d At the top of the Symbology pane for the Multivariate Proportional Symbol layer, click the
Vary Symbology By Attribute button .
g For Color Scheme, click the down arrow and check the box for Show Names.
This color scheme has been added to the project for you to make symbolizing easier. It is a
discrete color scheme that simply matches color stops to the codes used for each political
party.
Now you have a proportional-symbol map with size related to total number of votes for the
winning party and color representing the party itself.
The addition of more sensible colors really helps to communicate the patterns more
meaningfully.
mute sparsely populated constituencies in Scotland a bit, but when viewed at a small scale
(zoomed out), those smaller constituencies in London are still hard to see by comparison. And
you have seen how using differently sized symbols (graduated or proportional symbols rather
than choropleth fills) helps overcome some of those drawbacks—though large symbol
overlaps can also compromise the map.
Another alternative thematic mapping technique is dot density, which uses small dots to
represent some value of the data to create a patterned fill. Multiple dots are used to reflect
the total value.
a In the Contents pane, make a copy of the Winning Share Of Vote - Choropleth layer that
you made earlier and paste it in the UK Election 2015 map item.
c Turn off all layers except the Dot Density layer and the Dark Gray Canvas basemap layer.
d In the Contents pane, ensure that the Dot Density layer is selected.
h Try some other values for both the dot size and dot value.
A good dot-density map has a balance between a dot size that is visible and a dot value that
creates a map where the dense areas have coalescing dots but which are not simply a large
splotch.
You will also remove the background line from the area because it gets in the way of this sort
of map.
j On the Properties tab, set the Outline Color to No Color, make the Outline Width 0 pt,
and click Apply.
This update improves the overall appearance. Without the administrative borders, the map
takes on a smoother appearance.
b Copy the Dot Density layer and paste it into the Multivariate Dot Density group layer.
c Paste four more copies of the Dot Density layer into the Multivariate Dot Density group
layer so that you have a total of five layers.
e In the Multivariate Dot Density group layer, rename the first Dot Density layer
Conservatives.
Next, you will use a definition query to set the dot colors to the political party colors.
f Double-click the Conservatives layer to open the layer properties and click the Definition
Query tab.
g Click New Definition Query and build the following clause: Where First Is Equal To
CON.
i In the Symbology pane for the Conservatives layer, in the Symbol column, click the box
next to First_Votes to open the Format Point Symbol pane.
j Change the dot's color to the Conservative blue and click Apply.
Note: Because there are so many dots on the map, it may take some time for the map to
redraw.
Now you have a map with only the partial dataset shown.
k Repeat the previous steps for the second Dot Density layer, using the following
information:
l Use the same approach to specify definition queries and symbol colors for two additional
layers—one each for the SNP (Scottish National Party) and LD (Liberal Democrats) parties.
m Finally, for the fifth Dot Density layer, specify a definition query that has four clauses:
This captures all other constituencies that have not been shown by the other layers in this
group.
n Rename the fifth Dot Density layer Other, and then symbolize it with the light gray
Independent/Other color.
Now you have an overall effect where dot density is colored by the winning political party.
You could create a separate layer for every party if you wish, though sometimes a map is far
clearer if you remove some of the detail. Here, you are showing the predominant parties and
removing a bit of visual noise by grouping every other constituency as Other.
q If you are continuing to the stretch goal, leave ArcGIS Pro open; otherwise, exit ArcGIS
Pro.
Conclusion
You have designed a range of different thematic maps that each show the data in different
ways. Along the way, you have learned about the differences between area-based thematic
maps and point-based thematic maps. You have applied some of the ideas that were covered
in Section 2 Exercise 2 for classifying data, and you have explored ways to design and modify
symbology.
There are other types of thematic maps that you can create in ArcGIS Pro. To explore some
additional thematic maps, click the Final Maps map tab. Examine each of the thematic maps
and experiment with their settings.
Ultimately, there is never a perfect thematic map, but making sure that your data
classification, map type, and symbology work in harmony will help make your choices make
sense for your map reader.
The map is covered in apparently randomly positioned paint splats. Actually, there are four
paint splats per constituency. They represent the winning political party, the runners-up, the
third-place party, and a final symbol that represents the remaining candidates. But what else is
going on? Symbols are different sizes, so they are proportional to reflect the number of votes.
They are different colors, so they represent the political party colors. The symbols are
randomly rotated to create a more organic appearance. There is quite a lot going on here, so
you will slowly work your way through how it is constructed.
The tasks in this stretch goal are intentionally high-level. Use what you have learned in this
exercise to create your Jackson Pollock-inspired map. Pour yourself a cup of coffee or
tea—this will take a bit of effort.
a In the UK Election 2015 map view, in the Contents pane, turn off all layers except the
Dark Gray Canvas basemap.
Again, you will use a point feature class of the election data to create a group layer.
b Right-click the UK Election 2015 map item and create a new group layer named Jackson
Pollock.
c Make a copy of the original Constituency Points layer, and then in the Jackson Pollock
group layer, paste four copies.
d Rename the Constituency Points layers, starting with the first layer, as follows:
• Others
• Third place
• Second place
• First place
The order is important because you are going to have four layers showing at once, and you
want the larger symbols at the bottom of the stack (the winners) with the others (second place
and so on) atop. If you positioned these in the reverse order, the larger symbols would
occlude those underneath.
a In the Contents pane, for the First Place layer, open the Symbology pane.
b Update the Primary Symbology to Unique Values and for Field 1, choose First.
You will assign picture marker symbols from the preloaded electionsplat style (located in the
Format Point Symbol pane under Gallery) to each point symbol so that each party has the
correctly colored paint splat. The paint splats will be quite large initially; you will scale them
after they are loaded.
c Near the bottom of the Symbology pane, for each symbol in the grid, click the point
symbol, and then in the Format Point Symbol pane, on the Gallery tab, using the same
colors as before, choose the appropriate paint splat symbol for that value from the
electionsplat collection.
Note: On the Gallery tab, click once on the paint splat symbol to choose it, and then click the
Back arrow to return to the Symbology pane. Because the initial paint splats are quite
large, you may need to scroll down in the grid at the bottom of the Symbology pane.
GRN Green
IND Independent
LAB Labour
LD Liberal Democrats
Other Other
PC Plaid Cymru
SF Sinn Fein
Now you will vary the symbol size by an attribute to turn them into proportionally scaled paint
splats.
f Expand Size.
You will now define a size based on the square root of the total votes received by the winning
candidate (which is how proportional symbols are calculated to ensure that the area of a
symbol relates to the value being mapped).
g To the right of the Field drop-down list, click the Set An Expression button .
h In the Expression Builder dialog box, ensure that Language is set to Arcade.
n In the Contents pane, turn on the First Place layer and zoom in to see the symbols on the
map.
You will notice a problem. All the symbols are on the large side, so you can add a constant to
the equation to scale them all larger or smaller.
p Click OK.
You now have a map of the winning parties for each constituency with a proportionally scaled
paint splat symbol. This looks good, but the symbols all look a little regular and overlap one
another.
a To add rotation to the symbols, in the Symbology pane, ensure that you are still on the
Vary Symbology By Attribute tab.
You have changed transparency at the symbol level, but you can also change the transparency
of the whole layer.
f In the Contents pane, ensure that the First Place layer is selected.
g On the ribbon, from the Appearance tab, in the Effects group, in the Transparency field
type 14.0%.
i In the Contents pane, in the Jackson Pollock layer group, right-click the First Place layer,
point to Sharing, and choose Save As Layer File.
j Browse to the same location that you saved your exercise data file, name the layer
Pollock_First, and then click Save.
The symbology specifications that you just set for the First Place layer are saved to a reusable
file.
k Open the Symbology pane for the Second Place layer, choose Unique Values, and then
specify Second as the field to map.
l At the top of the Symbology pane, click the Menu button and choose Import
Symbology.
The Apply Symbology From Layer tool opens in the Geoprocessing pane.
m In the Geoprocessing pane, for Symbology Layer, click the Browse button .
n Browse to and select the Pollock_First layer file that you saved and click OK.
Note: You can also use an existing layer.
o In the Geoprocessing pane, under Symbology Fields, verify that Type is set to Value Field,
and then set Source Field and Target Field to Second.
p For the Transparency Expression Field section, set Source Field and Target Field to
Second_Share.
q Click Run.
At the bottom of the Geoprocessing pane, you should see a successfully completed message.
If you do not, it is likely that you entered something slightly wrong, so go back and check your
work.
r Repeat the process to apply the symbology from the layer file for the Third Place layer,
again making the symbols a little smaller and more transparent.
s For the final layer (Others), use the Symbology pane to create a single symbol, using a
gray paint splat for Other.
Note: Do not forget to scale the gray paint splat.
t In the same way that you applied 14.0% transparency to the First Place layer, change the
overall transparency of the Second Place layer to 39.0%, the Third Place layer to 20.0%,
and the Others layer to 46.0%.
You should end up with a Jackson Pollock-style mix of symbols that quite literally paints a
picture of the mess of the election.