PLSI 3346

Geography & World Politics

Day Two: Map Making

GIS is a powerful tool for displaying spatial data. Creating clear, informative maps that are true to the underlying data is something the software can do well. However, it is just as easy to create maps that are hard to understand—or worse—misrepresent the data you are trying to display.

Ideally, a good map should tell a story.

Let’s attempt to illustrate how greenhouse gases compare with oil consumption for the countries of the world.

à Symbology and Data Display

First, let’s figure out where to get our data.

Which table/s contain the right fields or attributes to answer the question of how greenhouse gases compare with oil consumption by country?

What are the fields called? What units are they in?

How must we operationalize this data in order to make a meaningful map-based display?

What is the range of data for each of these two fields?

How best can this range be classified?

A common way to display ‘count’ data (ratio data) on a map is to make what’s called a choropleth map.

This displays data from a single attribute classified by color.

Let’s start with greenhouse gases. Since there are three fields that contain greenhouse gas data (Total greenhouse gas emissions, Total greenhouse gas emissions per capita, Petroleum greenhouse gas emissions), let’s choose one to start with.

If we want to make a choropleth map of Total greenhouse gas emissions by country, where do we start?

First, let’s examine the data.

What is the range of data values?

What does ‘0’ mean in this data set?

How might we group, or classify, sets of data values?

ArcMap lets us do this kind of analysis very easily. We could open the attribute table and look at the column, and use the Sort and Statistics tools to analyze the data.

A more robust way of examining the distribution of data values can be found in the Symbology tool.

By right-clicking on the file name and selecting ‘Properties’ from the menu, you can open the Symbology dialogue box:

Now we can examine the data distribution, and break the range of data into classes, which we can then assign to a color scheme.

After choosing the appropriate field, click the Classify button (circled above).

This lets you see where your falls along the entire range, and lets you create classes.

Let’s take a minute to try some different kinds of classification.

What classification scheme (number of classes, breaks) works best for this unit, range, and distribution of data?

Once we’ve decided on classes, assigning a color ramp is easy.

Pick something that makes understanding the data easy on the eye. We naturally associate darker colors with ‘more’ or denser values.

Choropleth maps are great at showing the data for a single attribute, but what about creating a display of multiple attributes?

The ‘Charts’ Symbology (circled above) will let us show data from different fields in comparison.

We must still consider the range and distribution of our data here, too.

Let’s take a few minutes to experiment with different kinds of charts.

Which charts are best at showing data based on different units?

What would be a great use of a pie chart? What would be a bad one?

Data Exclusion:

Often, a map can be too busy with graphics if we symbolize the entire range of data. Data exclusion allows us to construct a formula to exclude any values or combinations of values our display doesn’t require.

This can make for a less cluttered and more informative map.

Often a good data display involves using colors and symbols in combination, usually from different layers. ‘Oil Consumption Graduated’ and ‘Oil by Production by Class’, when viewed simultaneously, are an example of this.