Tufts GIS Center

Cartography Tips
Author: Barbara Parmenter, Revised 9/5/2011

Note: for detailed help on making scale bars and legends in ArcGIS, see the tip sheet for Creating
and Editing Scale Bars and Legends on the Tufts Tip Sheets wiki under the Mapping and
Presentation section.

Cartographic Requirements for Project Maps

All maps must have the following:

• A title
• A scale bar or scale text showing map units (e.g., miles, meters)
• A north arrow
• A legend (quickly understandable by viewer - real words, not file names (e.g., “water” or
“lakes and rivers”, not Hydro_25k)!
• Source citations, including source agency or agencies, time period of content.
• Your name as cartographer
• Map projection used (e.g., Massachusetts State Plane Mainland, NAD 1983, meters)

Note that the spatial granularity of elements like census geography or parcels or county
boundaries must also be clearly noted either in the title, subtitle, or legend (e.g., population
density by census tract, land use by parcel)

General principles
• Always keep your map viewer in mind! Lay out your maps, text and other material to
help the viewer quickly understand what you want to show and to navigate through
your narrative. Never make the viewer work too hard!
• Provide both context and focus.
o The context is “where are we” in terms of a regional or metropolitan
perspective. This is usually accomplished by a map showing the study area in its
larger geographic region. It may also be accomplished at a more local level by
showing major roads and water in pale colors in the background.
o The focus is where you want the viewer to concentrate. This can be
accomplished in many ways – larger maps of the study area; the study area in
fully saturated colors (e.g., clipping the land use focus area and showing that in
full saturation but maintaining the wider land use context in the same colors but
at lower saturation. Note: the only way I know of to lessen saturation of a layer
is to increase its transparency.
o Don’t include layers or information that are unnecessary for your main focus.
• Be sure your viewer can distinguish categories and classes
1
Tufts GIS Center

o In maps showing ranges of numbers or categories, include as few classes as

possible to convey your information. Viewers can only comprehend about 5-7
categories and color ranges.
o In maps showing ranges of numeric values stick with one color ramp. If you are
showing hot and cold, or above/below mean, you might use a bi-color ramp like
blue to red but take great care with these kinds of color schemes.
o Use standard colors where appropriate (e.g., see the American Planning
Association’s recommendations for standard land use colors at
https://wikis.uit.tufts.edu/confluence/download/attachments/38341714/APA_
RLBCSColorForAllDimensions.pdf)
• Use white space effectively – make your map as large
as possible for the size space you have, then place
map elements (e.g., title, scale, legend, etc.) in the
remaining white space. The census map of Travis
County, Texas, at right, is an example.
• Lower the clutter quotient! Don’t repeat things in the
title, legend title, and legend classifications. Don’t
have unnecessary decimals in your tables or legends.
Don’t repeat units or % in your tables or legends. Use
these elements effectively to get information across without repetition. E.g, a title might
be “Transit Ridership in Boston” with a subtitle “By Census Tract (2000 Census)”. The
legend title might be “Percent of Workers Using Transit”, and the legend classification
could have <10, 10-30, 30-60, >60)

Setting the Data Frame’s Coordinate System

If the data you are using is already projected (e.g., data from MassGIS is already projected into
the Massachusetts State Plane NAD 83 coordinate system), you most likely will NOT want to
change the coordinate system of your data frame. However, if the data is unprojected (in a
geographic coordinate system – GCS), then you should project it for better map results. There
are many national data sets that are unprojected, including the US data we have from ESRI. The
example below shows how to project a map US data into an appropriate coordinate system.

US Counties – Geographic Coordinate System, US Counties – US Contiguous Albers Equal

NAD 83 (not projected) Area (projected)

2
Tufts GIS Center

To change the coordinate system of your map, e.g., from what you see on the left of the
previous image to what you see on the right (a much better way to show the US):

1. Choose View – Data Frame Properties (this

takes you to the properties of the entire
data frame)
2. Click on the Coordinate System tab
3. Select Predefined – Projected Coordinate
System and the type of the coordinate
system you want. In the case above we
went to Continental – North America – USA
– Contiguous Albers Equal Area Conic USGS.
4. Click OK when done.

Map Tips
• Don't use fill patterns (hatching, dots, etc.) - they are usually very distracting and hard to
distinguish – use them if you are forced to because you need to have a black and white
map. There ARE a few other occasions where they come in useful, but be careful and
sparing with them.
• Focus your viewer on what you want to show – include background / context data in
paler colors (e.g., roads in gray, water in a pale blue)
• For choropleth maps, is simplest to use a single color scheme (e.g., pale blue to dark
blue, white-gray-black, pale red to dark red) - there are some other color schemes that
work but the safest bet is a very simple color scheme.
• Use the Natural Breaks method of classification unless you think another method more
effectively tells the story you want to tell and understand the reasons behind your
choices. But do adjust the breaks to make more sense to a map viewer (e.g., 13.9%
might be better set at 15%).
• When you are first starting with mapping in ArcGIS, stick with the graduated colors
option, again unless you see a pressing need for something else. Other options like
graduated symbols or dot density can sometimes be appropriate, but you should wait to
use them until you have a better understanding of what impact they convey. Take a look
at using charts, like pie charts (e.g., you can show in a pie chart the distribution of
vacant, owner-occupied, and renter-occupied units), but again, unless you feel very
confident of your design skills, hold off for now. These make good maps in very rare
circumstances. Just because the computer can do it, doesn't mean you should use it!

3
Tufts GIS Center

• For some layers, there are too many distinct features to show each one by a color (e.g.,
zoning or watersheds). For these it is often better to simply have the feature outline
appear with no fill, and use the name or some other attribute to label each feature.
• Softer, pastel colors are more pleasing to the eye than bright colors (e.g., you can use a
pale versions of the standard land use colors for a very effective land use map)
• For background on your different topic maps, you need to include orientation features
like streets and rivers but in such a way that they don't distract from the main topic (e.g,
land use) - you can color streets a light gray so that the viewer can see them but they
are not so dark as to overwhelm the map.With major road and water, also be discrete.
• If your outline colors on some layers are getting in the way, turn them off with the
symbology properties or work to adjust them so that they don't distract from the main
idea of the map.
• The study area shape may be a challenge to show on paper. To get the most for your
space on the actual map, make the data frame on the layout fit your study area, and
then put the excess white space on the paper to good use for your other map elements.
• Make sure your scale is easily readable and makes sense - adjust the size so that it is
reading whole miles - e.g., 2 miles, not 1.8 miles.
• Adjust the number of columns on the legend so that you can spread it across space if
you need to
• Make a digital version of your map before printing on paper or a plotter – you will see
how the various elements come out looking on the page size you specify – adjust as
necessary

Useful functions
To constrain a map to a certain area (e.g., town boundary, pedshed) enable “virtual clipping” by
a shape file - Data frame properties – Data Frame Tab – Enable Clip to Shape. But use this
sparingly – you are eliminating the context when you do this. It could be appropriate in the
context of a poster where another map or maps clearly indicate context.

The Data Frame toolbar has a tool that allows you to rotate the map and north arrow for a map
of an area that does not fit well into your standard rectangle framework (e.g., Somerville).
Again, use sparingly – people are used to map orientation where up is north, so it is easy to
confuse folks where north is pointing to some other direction.

Get rid of outlines where appropriate – e.g., if outlines are getting in the way of you seeing the
color differences between parcel land use or between census tract population density.

If you have a good coloring and classification set up for a layer, save it as a layer file. Then you
can apply it to a clipped copy easily.