Why county maps can fool your eye

On US County Area, each outline is fitted independently into a 640 × 360 coordinate frame. The fitting process uses the available width and height for that county. A compact county, a long narrow one, and a scattered island geography can therefore use the frame differently. The apparent size of the filled shape is partly a consequence of its proportions.

That design is useful for recognizing a place at small screen sizes, but it removes a common scale between separate maps. You cannot compare their pixel areas as if a pixel represented the same number of square miles in both. Read the numeric measurements or open the comparison tool, whose land bars use a shared scale for the selected pair.

Suppose you represent a 100-square-mile area with a square. Its side is 10 miles. A square representing 10,000 square miles has a side of 100 miles. The second square has one hundred times the area but only ten times the side length. This square-root relationship matters when drawing proportional symbols.

Using real measurements, San Bernardino County, California has 9,808.6 times the land area of Falls Church city, Virginia. Equal-area comparison squares would have side lengths in a ratio of approximately 99 to one. These are conceptual squares, not their actual shapes and not a statement that one outline can be packed into another. Compare their recorded measurements before relying on the visual impression.

A projection translates Earth’s curved surface to a flat map. USGS explains that projections preserve different properties and cannot preserve area, shape, distance, and direction everywhere at once. An equal-area projection is designed for area relationships; it does not guarantee that every shape or distance is preserved. Read the USGS overview.

Even an equal-area projection does not rescue a comparison if you resize each county independently afterward. Projection controls how geography becomes planar geometry; fitting controls how that geometry occupies the screen. Both need attention. Our small outlines use a locally centered equal-area projection and then independent fitting, so they remain unsuitable for comparing screen area across different county pages.

A tiny map with every boundary vertex can be slow and visually cluttered. Generalization removes detail to produce a more readable illustration. Census cartographic boundary files are intended for that kind of small-scale thematic mapping, and their documentation explicitly excludes area and perimeter calculation from the intended uses.

Our illustrated outlines use the 2025 cartographic boundary release. The county measurements come separately from the 2026 Gazetteer. We do not calculate those measurements from the SVGs. This distinction also explains why tracing a coastline in the illustration cannot reproduce the water-area number: simplification and offshore clipping change what is drawn. Use the source limitations when selecting boundaries for a new analysis.

For “which county has more land?”, a labeled bar chart with a common zero baseline is often clearer than two outlines. For “what shape is the county?”, the outline is appropriate. For “where is it relative to neighboring places?”, use a contextual map with surrounding geography. For “what share is water?”, use a labeled numerical breakdown or a proportional composition bar.

A map can answer more than one question, but it needs the appropriate construction and labels. If you publish a shared-scale outline comparison, preserve the same scale through export and layout, explain the projection, and identify whether you display total boundaries or a land-focused shape. Do not claim a precise scale comparison after independently resizing the finished images to fill matching cards.

1. Look for the scale. Determine whether it is common to every map or local to each panel.
2. Read the measurement label. Land, water, and total area are different quantities.
3. Check the boundary source. A generalized illustration is not a survey or a measurement dataset.
4. Check the vintage. A shape and a table may describe different releases.
5. Test one numerical pair. If the picture suggests equal size while the recorded areas differ substantially, investigate the rendering before interpreting the geography.

These checks also help when designing your own chart. Include the caveat where a reader sees the map, not only in a distant methods page. A small note that outlines are independently fitted can prevent a confident but incorrect comparison without making the graphic harder to use.