We all have a story to tell and maps can help visualize huge amounts of data into a gorgeous, simple image. But how do you know which symbol to use and what do you call that dang map with bubbles? I mean circles, dots, points, or what the heck - let me just grab a piece of paper and draw it!
Well here you will learn what those lovely symbols are called and how to engage in the best dinner party conversation with something like:
"Did you see that compelling map visualization in the New York Times yesterday?"
"Why yes, I just love the way they overlaid a dot-density over a choropleth map to show the impacts of global warming"
People will be begging for more...
So where to begin... how about with the person reading the map?
The average map reader will scan a map for six seconds in search of its main message. You had better be pretty deliberate in your choice of symbology to tell your story quickly and effectively.
In this Story Map, we will help you make a great first impression by choosing the most effective symbology. We will wrestle with:
how to use size, shape, orientation, color hue, value (lightness), texture, color saturation, transparency
how our brains perceive visual data
how data type affects color scheme
how to classify data without misleading
how to simplify crowded features with averages, prevalence, transparency and glow.
This Story Map will present a menu of symbology to choose from and offer some tips (💡) on how to use each and some sparkle (✨) to celebrate what is great about each type of symbology.
Before whipping up a map, cartographers and communicators must first look deep down and ask some tough questions like:
What is the story you are trying to tell?
What is the intended message?
Who is the intended audience?
What do they think about the topic?
What do I want them to realize?
What data do I need to tell this story?
Cartographers craft maps by selecting the information that matters most and determining how to best present it. The resulting maps selectively show what we want people to care about, so they have the power to introduce bias, distort and lie. Gasp of horror.
For more on how to lie with maps, read Mark Monmonier's How to Lie with Maps.
Data drives symbology. Make sure to pull back the curtain and look into your data. This is pretty straight-forward:
Data can be numbers or words.
For numerical data, the number can be continuous with infinite options like weight with 1/4 pound increments or discrete with finite options like number of people.
For categorical data, the words may suggest an order (ordinal) such as small, medium, large - which has a hierarchy - or name (nominal) such as Girl Scout cookie types, eye color, or dog breed - which has no hierarchy.
Your symbology will differ based on the type of data. For example, Numerical data are often shown as a change in symbol size or intensity of color. Nominal data are often shown as a change in symbol shape or color hue.
So let's dive in and celebrate the thematic map as the powerful rock-star communication tool it is for all things point, line, and polygon!
If you need to show basic location and the data is distributed geographically, use points.
This map shows the capital city and capitol building for each state.
💡The power of this map is in its simple stars. Even without a legend, most people associate stars with capital city. So use a point symbol that is familiar to your audience.
✨Simple point maps are satisfying as they show location and distribution.
If you need to show basic location and the data are clustered geographically, then you may need to simplify the symbology.
Here are the point locations of major United States cities. They are too crowded.
You can make the points smaller. But in this case they still overlap.
You can even use color to discriminate between two different sized cities, but it still looks crowded.
One solution for overlapping points is to highlight them with a firefly effect. This uses size and transparency to make the points appear to glow against a dark background.
💡Firefly effects really pop against a dark background. It can be used to imply that a resource - like a lighthouse on a stormy night.
This map shows major cities by population class (discrete data) with the larger, brighter symbols showing the cities with the higher population class.
Download the firefly style pack âž
✨Firefly symbology is great for illuminating categorical and numerical data with light, bright halos of different color hue and size.
If your points represent numerical data, you can classify cities into small, medium, and large using graduated symbols.
Anytime you classify your data, you need to choose the optimal number of classes and where to put the breaks between those classes (breakpoints).
Here, a high transparency is used, so you can see other smaller cities below larger ones.
The size of the symbol does not necessary indicate the size of the value. So the symbol for a city with 5 million people is not 10 times as large as a city with 500,000 people.
✨The beauty of the graduated symbol is that it allows you to rank into simple buckets like small-medium-large, which can be easier for humans to discriminate between than seeing many sizes.
If you want the map reader to be able to better gauge the absolute (rather than just relative) values, use a proportional symbol map since the size of the symbol is proportional to the value of the data.
Here symbol classes are rounded to the closest million.
💡To create proportional symbols, you need to be sure that your break points are at equal intervals and that your minimum and maximum point size reflects the range of values in your data.
✨Proportional symbols are perfect for accurately reporting a data value because the symbol area is proportional to the data value. It is really intuitive.
Alas, New York City is the biggest U.S. city.
If your point data is clustered closely together in some locations, you can try a heat map to help see the density of the locations more clearly.
In a heat map, the point data are analyzed to create an interpolated surface showing the density of occurrence.
Here is a heat map of major cities.
✨Heat maps are fantastic at creating an index out of a given set of numbers or locations. It gives you an instant snapshot of the density of your data.
Aha - there are lots of major cities in southern California and the northeast!
Another way to represent many points is to cluster the individual point symbols into in a larger sized point that reflects the average value of all the features in that cluster.
This map shows the average values for median age of people living in major U.S. cities. Note that the clustering changes as you zoom in.
The size of the cluster symbol is scaled based on the number of features represented.
The color shows younger or older.
✨Cluster symbols can be valuable for exploring the trends of multiple variables in your data.
✨They are also effective at aggregating your data if it is crowded geographically or if you have a variety of values are want to see the average value.
Ahoy, there are lots of older people in Florida and lots of younger people in L.A.!
These multi-colored lines show the location of transmission lines. Unique colors help discriminate between each type.
Here they also used firefly points to show power plants. Great association between power and light!
Web map by Jeffrey McDonald, Energy Information Administration (EIA) âž
💡Remember that colors have implied meaning, so map readers will associate symbology colors with the appearance of the item as in this satellite imagery.
blue with water
green as park or natural
grey or brown for cities
Some lines do not show actual physical features like roads or utilities. Instead these isolines lines show where a value is the same.
iso = same
line = line :-)
If you have data sampled a random or regular intervals, isolines can make it easier to visualize this over an area.
On this map, isolines are used to show the line with the same depth (bathymetric map).
💡Darker lines often indicate a deeper depths. Darker saturation of colors often indicate a greater value.
Isolines can be also used to show the same elevation (contour lines in a topographic map, pressure (isobars) or temperature (isotherms).
✨Isolines are great for creating a continuous surface when you have data sample for a set of irregularly spaced point locations like weather stations.
If direction is important in your data for flow like traffic, current, then use line symbols to show that direction.
Red is used for hot and blue for cold just like water taps. It is a universal association of color.
If your data involves an origin and destination, try a flow map.
Flow maps use line thickness to show the proportion of traffic or flow coming from different places.
This map show the origin cities of conference attendees.
Orange is the complimentary color (opposite on the color wheel) of blue, so the contrast is great over the ocean.
✨Flow maps are the perfect choice when you want to show movement and amount of "traffic" flow.
Polygon data isn't always obvious. Sometimes it can be symbolized using a single point or many dots. In the end, these examples are all polygons - that is a feature with an area.
Sometimes a label can act like a symbol.
This map uses a color-coded label to show value for each polygon. It uses the north-south and east west center point or centroid for the label location.
This map shows happiness rankings for each country.
Notice how the map is colored:
blue (happy sky color) for happy
orange (warning color) for medium
red (danger color) for unhappy.
Another useful color scheme for ordinal data is the stop light set of:
red (stop)
yellow (warning)
green (go)
This color scheme can also be used to signify bad and good.
✨Colored centroid labels are an effective way to convey detailed quantitative data by using color to show the overall trend or pattern.
Human eyes perceive trends in color hue - much easier than shape.
If you have numerical data for different geographic areas and want to compare them, a proportionate or graduated symbol map could work.
A graduated symbol map has small, medium, and large symbols, but there is no guarantee that the symbols are proportionate to the data.
A proportional symbol or dot distribution map scales the size of the symbol proportionately to a data value. So an area that has twice the value will be twice as large.
This maps show the number of renters in each county.
You can even use a proportionate symbol map to show multiple variables.
Here both renters and those without health insurance are shown.
✨Proportional symbols are powerful way to distill and visualize magnitude or count data.
Choropleth maps use color or shading to show a value or category.
Easy to remember because:
"choro" = color
"pleth" = excessive
Woohoo - cartographers really know how to be excessive!
If you have categorical data or standardized numerical data, you can use a choropleth map to compare areas.
Standardized numerical data means that the data is in rates or ratios (% of people) that allows you to compare areas -- rather than raw count data (number of people) that may be skewed depending on the size of the area.
By standardizing, you are comparing apples to apples.
This map shows obesity rates as % of the population.
Remember that for choropleth maps you have to classify your data and choose a color scheme based on the data type.
More color schemes and when to use nominal, sequential, or diverging âž
✨Choropleths are valuable for comparing categories or rates/ratios across different geographic areas.
If you want to highlight the category that occurs most frequently, you can use a predominant category map.
This map shows the predominant educational attainment in a given census tract.
These education data are ordinal because there is a hierarchy of amount of education. The cartographer used unique hues but provided order by ranking the hues by wavelength. The shortest wavelength is the greatest education.
Even the rainbow is recognized as an intuitive order.
✨The magic of predominant category maps is that it evaluates multiple fields of data and tabulates the predominant category for you.
Predominant maps are also great for highlighting trends in the categorical (word) data. This is because categorical data cannot be standardized into a rate or ratio like numerical data.
If you want to show density differences across geography without the potential bias of a choropleth with its break points between each value, then use a dot density map.
A dot-density map uses dots or other symbols on the map to show the values of one or more numeric data fields. Each dot on a dot-density map represents some amount of data.
In this case, one dot = two households.
💡Ideally you want to use an equal area projection, so that all areas are in proportion to the dots. So that the map reader sees an accurate view.
Blue is used for the greater income households. Red is the at-risk population with low incomes.
✨Dot density maps show density differences across geography as a pattern. No break points or bias are involved.
What you see is what you get!
Hexagons, and other regularly shaped features like squares/rasters, allow you to normalize geography rather than being limited to using irregularly shaped polygons such as county boundaries, census tracts, zip codes.
This map show predominant land use.
✨Hexagon maps are great for assessing areas as they are free from the bias that an irregular shaped area can have.
A cartogram distorts the size of geographic area proportionally to the value. This one is called a non-contiguous cartogram because the sides of each state do not touch each other.
Here the size of each state reflects the number of votes. Blue denotes the traditional color association of Democratic and red denotes Republican.
✨Cartograms are really easy for human eyes to take in because they use size to represent data.
This map type distorts size to show numerical data and color to show nominal data.
💡Cartograms are especially valuable when there is a significant variation in the ratio between your count data and the actual geographic size.
If you want to know the half-way point between two locations, in other words what is the closest resource, a Voronoi Diagram is what you need.
It looks at all of the point locations and creates lines dividing the surrounding area half way between each point.
It is good for determining an area served as the crow flies - so to speak (aka Euclidean distance). Here kindergarten locations are analyzed to determine the closest kindergarten for each area.
💡To cross language barriers, use a literal (rather than arbitrary) symbol, as is done here with schools. Universally understood symbols don't need a legend and you won't need to know how to speak German!
Bonus points if you know how to say "kindergarten" in German.
✨Voronoi Diagrams are fantastic for showing the nearest resource, planning the distribution of resources to be as far away from other resources.
To determine how far you can drive within a certain period of time, use a drive-time analysis.
This drive time maps shows the areas where people can drive to a legal service office within 60-minutes.
I sometimes call these "octopus" maps to explain what they look like because the road corridors often reveal far-reaching legs from the center point.
💡Research and select a drive time that is reasonable given the traveler and the destination. People will travel a short distance to get coffee but will travel large distance to get neurosurgery.
✨Drive-time maps are a practical way to show the distribution and availability of a resource.
The size and shape of the drive time area varies based on time of day and because of access constraints such as water features, mountains, lack of roads or high-speed road network, and traffic.
Speaking of time, there are some great ways to show time and age when mapping point, line, and polygon data, but that could be a whole other story map or two or three. Check out this story map on how to show time in maps âž
And our time is up!
Hopefully these examples have whet your appetite and provided food for thought. Have fun getting creative about how to best symbolize your data for the greatest impact. Visit the resources below for inspiration and information!
Best, Alison
An error has occurred |