Logarithmic scales represent equal ratios with equal distances. On an ordinary linear scale, equally spaced marks might show 10, 20 and 30. On a base-ten logarithmic scale, they might show 10, 100 and 1,000. The difference matters because the second scale shows multiplication rather than equal additions. Scientists use logarithmic scales when values span a very wide range or when proportional change is more informative than absolute change. You will encounter them in earthquake reporting, sound measurements and graphs of growth. Learning to read the axis carefully prevents a small-looking gap from disguising a large numerical difference. It also helps explain why a straight line can sometimes represent accelerating growth rather than a steady addition.
Key takeaways
- Equal distances on a logarithmic axis represent equal ratios, not equal numerical differences.
- A base-ten step multiplies a value by ten; it does not add ten.
- Always check whether one axis, both axes or neither axis is logarithmic.
- A logarithmic graph can reveal proportional patterns while compressing large absolute differences.
- Earthquake magnitude and decibels require attention to what physical quantity is being compared.
How logarithmic scales work
A logarithm answers an exponent question: what power of a chosen base gives a particular number? For base ten, the logarithm of 10 is 1, of 100 is 2 and of 1,000 is 3. Those logarithms increase in equal steps even though the original numbers do not.
A logarithmic axis usually labels the original values rather than their logarithms. You may therefore see 10, 100 and 1,000 at equal distances. The positioning uses logarithms, while the labels tell you the quantity being represented.
Think of each equal interval as applying the same multiplier. Moving from 10 to 100 is a tenfold increase, just like moving from 100 to 1,000. The second change adds far more in absolute terms, but its ratio is identical.
Compare a linear and a logarithmic reading
Suppose three fictional samples contain 10, 100 and 1,000 particles. On a linear axis beginning at zero, the first two values sit relatively close together compared with the third. That graph emphasises the absolute differences in particle counts.
On a logarithmic axis, the three values are equally spaced because each is ten times its predecessor. This graph emphasises their proportional relationship. Neither representation changes the data; each makes a different feature easier to see.
The appropriate choice depends on the question. If you want to compare the extra particles between samples, the linear scale may be clearer. If you want to compare multiplication across several orders of magnitude, the logarithmic scale may be more useful.
What does an order of magnitude mean?
In a base-ten comparison, one order of magnitude corresponds to a factor of ten. Two orders correspond to a factor of one hundred. The phrase is often used approximately, but the underlying idea is a difference in scale rather than a small additive change.
A value of 1,000 is two orders of magnitude above 10 because 1,000 divided by 10 equals 100. It is not three orders above it simply because 1,000 contains three zeros. Always compare the two values rather than inspecting one number in isolation.
Read the spaces between the major marks
The midpoint between 10 and 100 on a logarithmic axis is not 55. Equal distances require equal ratios, so the midpoint is approximately 31.6: multiplying 10 by about 3.16 reaches 31.6, and multiplying again by the same factor reaches about 100.
This midpoint is the geometric mean of the two endpoint values, obtained by taking the square root of their product. The arithmetic mean, 55, belongs halfway between them on a linear scale instead.
Intermediate marks such as 20, 30 and 40 are therefore not evenly spaced. They crowd closer together as the numbers approach 100 within that decade. If you estimate a value by treating those spaces as equal additions, you will misread the graph.
Why scientists compress wide ranges
Imagine plotting fictional populations ranging from ten organisms to ten million. On one ordinary linear graph, the smaller values may be visually indistinguishable near the baseline. A logarithmic axis gives each tenfold interval comparable space.
This can reveal differences among small values without discarding the largest ones. It is useful when the question concerns relative scale across a broad range. However, the compression also makes enormous absolute gaps appear modest.
Responsible interpretation therefore reports actual values or ratios, not just visual impressions. Saying that two points are close together is incomplete unless readers know the scale. A small distance can represent a substantial difference in the measured quantity.
A straight line can show exponential growth
Consider an invented population that doubles every day: 10, 20, 40, 80 and 160. On a linear graph, the daily additions grow larger, so the line curves upwards. It represents constant proportional growth, not constant numerical growth.
On a graph with a linear time axis and a logarithmic population axis, the same values fall on a straight line. Each day applies the same multiplier, and equal multipliers occupy equal vertical distances.
This is called a semi-log graph because only one axis is logarithmic. The straightness suggests a consistent growth rate over the displayed interval. It does not demonstrate that the pattern will continue indefinitely, or identify the mechanism causing it.
What a changing slope tells you
If the line becomes less steep on a semi-log graph, proportional growth has slowed. The population can still be increasing rapidly in absolute numbers. A flattening curve does not necessarily mean that the quantity is falling.
For example, an increase from 100 to 150 adds 50 but represents 50 per cent growth. An increase from 1,000 to 1,100 adds 100 but represents only 10 per cent growth. Larger additions can coexist with a lower proportional growth rate.
When both axes are logarithmic
A log–log graph uses logarithmic scales on both axes. This is different from a semi-log graph and changes the interpretation of straight lines. A straight relationship on a log–log graph can indicate a power-law pattern over the range being examined.
As a simple mathematical example, the area of a square equals the square of its side length. Doubling the side multiplies the area by four. Plotting side length against area on logarithmic axes makes this power relationship straight.
Do not assume that every nearly straight cloud of points proves a universal law. Measurement uncertainty, a limited range and alternative models may all matter. Earth Site’s guide to measurement uncertainty and meaningful differences explains why an attractive pattern still needs careful evaluation.
Earthquake magnitude: separate size, energy and effects
Earthquake magnitudes use logarithmic relationships, which is why a one-unit difference is substantial. In the original local, or Richter, magnitude framework, a unit increase corresponds to a tenfold increase in the corrected recorded wave amplitude used by that scale.
Energy comparisons are different. The commonly used approximate relationship gives about 32 times more released energy for a one-unit magnitude increase, and about 1,000 times more for two units. Multiplying a factor twice is not the same as doubling it.
Modern reporting often uses moment magnitude, derived from seismic moment rather than simply the largest trace on a seismogram. Magnitude is also distinct from shaking intensity at a location. Distance, depth and local ground conditions influence the effects people experience. Earth Site’s introduction to the causes of earthquakes provides broader geological context.
Decibels: the reference and quantity matter
Decibels express a logarithmic level or ratio. For sound intensity, a ten-decibel increase corresponds to a tenfold intensity ratio. An increase of twenty decibels corresponds to a hundredfold ratio, provided the comparison uses the same relevant reference and conditions.
Sound pressure is related differently: decibel calculations for pressure amplitudes use a factor of twenty in the logarithmic expression, while power-like quantities use ten. This is why it is important to identify the quantity rather than memorise one multiplier for every situation.
Physical intensity is not identical to perceived loudness. Hearing depends on frequency and other conditions, so “ten times the intensity” should not be casually rewritten as “ten times as loud”. The numerical scale describes a defined physical comparison, not a simple universal measure of human experience.
Why ordinary logarithmic axes exclude zero
No finite power of ten equals zero. Consequently, zero has no finite logarithm on a standard real-valued logarithmic axis. Negative original values are also excluded from the usual logarithm used in these graphs.
This creates practical questions when a dataset contains zeros or changes sign. A graph may omit such values or use a different transformation. Adding a constant before taking logarithms changes the relationship being displayed and should be disclosed.
A negative decibel value is not a contradiction. It can indicate that a positive measured quantity is below its reference level. The logarithm of a positive ratio smaller than one is negative; the original physical quantity need not be negative.
A reliable graph-reading routine
Read the axis titles and units first, then inspect the numbered marks. A sequence such as 1, 10, 100 suggests a base-ten logarithmic scale. Check whether both axes use that structure and whether any breaks or special transformations are indicated.
Choose two points and calculate their ratio as well as their absolute difference. This turns visual impressions into explicit comparisons. Check the time span, sample coverage and uncertainty before interpreting a line’s shape.
Finally, distinguish description from explanation. A graph may show a regular pattern without establishing its cause. The broader discussion of how scientific evidence supports explanations helps connect careful graph reading with sound scientific reasoning.
Conclusion
Logarithmic scales are useful because they make multiplication visible. They allow wide ranges and proportional patterns to fit on readable graphs, but they change what distances and slopes mean. The safest habit is to inspect the labels, identify the measured quantity and translate visual gaps into numerical ratios. That small adjustment prevents major misunderstandings in science reporting and everyday data interpretation.
Frequently asked questions
Does a logarithmic graph change the data?
It changes their visual spacing, not the underlying values. The scale emphasises ratios and compresses large absolute differences. Clear labels are essential so readers know how to interpret the display.
Are all logarithmic scales base ten?
No. Other bases, including two and the mathematical constant e, are used. The defining property remains equal spacing for equal ratios. Read the labels or method description to identify the base.
Does a straight semi-log line mean constant additions?
No. With time on a linear axis and the measured quantity on a logarithmic axis, a straight line indicates a constant proportional rate over that interval, not equal numerical additions.
Can logarithmic scales be misleading?
Yes, if readers are not told about the scale or if compression hides differences relevant to the question. They can also clarify data that a linear graph makes unreadable. Suitability depends on purpose and labelling.
Why is an earthquake one magnitude larger not just slightly stronger?
Magnitude differences are logarithmic. A one-unit increase corresponds approximately to 32 times the released energy, although local shaking and damage depend on additional factors beyond magnitude alone.
