CEE 8813 · GIS Lecture 1 · Classed maps

Classed Map Lab

On a graduated color map, the map maker chooses the classes. This page maps all 159 Georgia counties from the lab's FEMA National Risk Index shapefile. Change the classification method, the number of classes or the normalization, and watch the same numbers tell a different story.

Hover or tap a county to read its value and class. Scroll or pinch to zoom, and drag to pan (with two fingers on a phone). Projection: Georgia Statewide Lambert style (Lambert Conformal Conic, standard parallels 31°25′N and 34°17′N).

Symbology

Method
5
Compare

Histogram of the 159 values

Bars are colored by class; the strip under them has one tick per county. Drag a break line to move it, which switches the method to Manual Interval, as in ArcGIS Pro's histogram view.

Legend

Each class includes its upper value, as in ArcGIS Pro. The number on the right is how many counties fall in the class.

Counties per class–smallest – largest
GVF–goodness of variance fit, 1 is perfect
No data0counties, hatched grey on the map

Compare two classifications

Same field, same normalization, same number of classes and colors. Only the method changes. Counties outlined in black land in a different class on the two maps.

Counties that change class–

What you are seeing

Try this in class

  1. Keep the NRI composite risk score and step through Equal Interval, Quantile and Natural Breaks. Watch the counts in the legend, and which counties make the top class each time.
  2. In the compare maps, find a county that is dark on the left and light on the right. Hover it and read its value in both classes. Then try Natural Breaks against Quantile.
  3. Choose Population with Normalization set to None, then switch Normalization to per sq mi. Fulton has the most people, but where does it rank once you divide by area? Then set Primary symbology to Proportional Symbols for the raw totals.
  4. Choose Expected annual loss and switch between None, per capita and per sq mi. Which counties fall out of the top class, and which rise into it?
  5. Drag the break lines on the histogram to round numbers such as 5, 10 and 15. Does your manual map support the same headline as Natural Breaks?
  6. Group exercise (slides 50–52): tell a story with the risk index. Each group picks one field, one normalization and one method, writes a one-sentence headline the map supports, then hands the controls to another group to see whether a different classification weakens it.

Map design checks before you export

Slides 36 and 53–56. Tick them as you check your own map; nothing is saved.

Back to Lecture 1: classed maps

Slides 42–52 move from what a classed map is, to how to classify the numbers, to the group exercise with the county risk index.

Classed maps and symbolsSlides 42–45

A graduated color map (choropleth) shades each polygon by class. A graduated symbol map puts a sized marker on each feature. Points take marker symbols, lines take line symbols and polygons take shaded fills. Interval and ratio data must be divided into classes before mapping, then shown by size, thickness or color.

Rule of thumb: totals (people, dollars) suit symbols, while rates, densities and scores suit a choropleth. In ArcGIS Pro: Symbology pane → Primary symbology → Graduated Colors, Graduated Symbols or Proportional Symbols.

Classifying data and layer symbologySlides 46 and 48–49

Same data, different classifications. The choice changes how the map looks and whether it tells the truth, and the best method depends on how the data are distributed and what the map is for.

ArcGIS Pro: right-click the layer → Symbology, then set Field, Normalization, Method, Classes and Color scheme. The histogram view lets you drag breaks, which switches the method to Manual Interval. The 2022 ArcMap slides use Layer Properties → Symbology → Quantities (slide 49).

Group exercise: the risk index by countySlides 50–52

Using the shapefiles you have (Georgia state-maintained routes and the county risk index), make a formal map and tell a short story with it. Read the data dictionary first and choose symbology that fits the field.

Watch out: the shapefile already carries FEMA's own national ratings (risk_ratng): 86 Georgia counties are “Very Low”, 58 “Relatively Low” and 15 “Relatively Moderate”. Those are national classes. The classes on this page are drawn from Georgia's 159 values only.

Pitfalls students hit

  • Mapping a raw total such as population or expected annual loss as a choropleth. Big counties look important because they are big. Normalize, or use symbols.
  • Using Quantile on skewed data. It forces equal class counts, so small differences in the middle get different colors while very different counties at the top share one.
  • Using Equal Interval on skewed data. A few extreme counties stretch the range, and most of the state ends up in one or two classes.
  • Comparing two maps (two years, two hazards) that each have their own Natural Breaks. Jenks breaks belong to one data set. Use the same manual breaks on both maps.
  • Keeping the default legend labels. Round the breaks and the labels, and state the units.