How to Read an Astronomy Research Paper Figure First

By Julia Berens · · 5 min read

A silhouetted figure atop a rock arch shines a blue beam into a starry night sky streaked with the pink and purple Milky Way.

Learning how to read an astronomy research paper does not mean understanding every sentence on the first pass. For Pre-Final Round practice, start with two things you can write on a sheet of paper: the question the authors are trying to answer and the number of a figure that might help answer it. Leave unfamiliar terms in the margin until you know which ones you actually need.

How to read an astronomy research paper without starting at page one

Read the title and skim the abstract for the research question. If the abstract leads with a result, turn that result into a question. For example, if the authors report that a galaxy appears wider in one wavelength band than in another, the question might be: How does the galaxy's measured width differ between the bands? If you cannot tell what is being compared, check the end of the introduction, where authors often state their aim.

Next, scan the figures and captions for the comparison you just wrote down. Some figures show where observations were taken or how measurements were checked. Those can matter later, but start with the plot that most directly addresses the question. If the paper compares two bands, look for a figure that shows both, either together or in matching panels.

Then read the paragraph in the Results section that refers to your figure. In a digital copy, search for the figure number as it appears in the text, such as “Fig. 2.” Pick one claim to check before reading the rest of the section. Save the detailed Methods reading for when you can name something specific you want to verify.

Read the caption with the plot in view

Before deciding what a plot shows, read the labels on both axes, including the units. Look at the tick labels, too: a logarithmic axis can put 1, 10, and 100 at equal intervals. Check the legend and caption to learn what the symbols stand for. A dot may represent one observation or an average of several. A line could connect measurements, show a model prediction, or mark a reference value. Its meaning comes from the paper, not its shape.

If there are several panels, identify the one you need by letter. The caption may explain which panels show different observations and which show the same data after different processing. Also check what any error bars or shaded bands represent. A bar showing uncertainty in a mean is not the same as a bar showing the spread of the individual measurements. If the caption does not define it, look in the Methods rather than guessing.

Try describing the plot in one sentence without explaining why the pattern occurs: “In panel b, the orange profile falls to half its peak farther from the center than the blue profile.” That sentence points to something another reader can check. An explanation of what makes the profiles different is a separate step.

Trace one claim back to the evidence

Imagine a practice paper with this claim: “The galaxy's light profile is wider in band A than in band B.” This is a made-up example, not a result from a published paper. Suppose Figure 2b places distance from the galaxy center on the horizontal axis, in arcseconds. The vertical axis shows brightness divided by the peak brightness in each band. The caption identifies the two curves and says the authors compare their widths using the radius at which each profile drops to half of its peak.

Start at 0.5 on the vertical axis. Move across to each curve, then down to the horizontal axis. Suppose you read about 4 arcseconds for band A and about 2 arcseconds for band B. You now have an evidence note you can use: “Figure 2b shows that band A reaches half its peak at about 4 arcseconds, compared with about 2 arcseconds for band B, using the caption's definition of the comparison.” The word “about” matters because you estimated the values from a plot.

That plot does not tell you which band has the brighter peak. Dividing each profile by its own peak makes both peaks equal to 1 on the graph. Nor does the difference by itself establish why the profiles are wider or narrower. If the authors say the difference reflects the galaxy itself, check whether they accounted for image blur in the two bands. If they claim the difference is larger than measurement uncertainty, look for uncertainty estimates on the radii or for a test that supports that statement. A visible gap is a place to start checking, not a substitute for that check.

Look up terms that could change your reading

Return to the words you marked on your first pass. Start with terms that define an axis, a quantity in the claim, or a step that could affect the comparison. In the example, “normalized brightness” matters immediately because it tells you what the vertical axis can and cannot show. “Point spread function” may matter next: it describes how an imaging system spreads light from a point source. If a point looks more spread out in one band, a galaxy profile in that band can also appear broader.

Check how the authors use a term before settling for a general definition. For instance, the word “width” by itself does not tell you whether the authors measured a radius from the center or a distance from one side of the profile to the other. The caption or Methods section should tell you how the quantity was measured. Write the meaning beside your evidence note, then reread the claim to see whether your interpretation has changed.

Practice writing an answer someone else can check

For Pre-Final Round preparation, try this exercise with an astronomy paper that has a plot you can read. The IAAC learning resources can help when you need background on the topic, and the guide to the competition rounds shows where the Pre-Final Round fits.

  1. Write the authors' question in one sentence. Include what they measured or compared.
  2. Copy a claim from the Results section and record the figure and panel it refers to.
  3. Write your own evidence sentence using the labels, units, and a value or pattern you can point to. Include a detail from the caption if it determines how to read the plot.
  4. Write one thing the figure does not establish. If the claim goes beyond the figure, look for another plot, a table, or a passage in the Methods that supports the extra step.

Give your evidence sentence and the figure to a classmate without showing them the authors' claim. Ask them to find the same feature. If they have to ask which curve you used or where you got a number, add that detail to your answer.

Frequently asked questions

Do I need to read the whole abstract first?

No. Skim it for the research question, then check the end of the introduction if you are still unsure what the authors are testing or comparing.

Can I estimate a number from a plot?

Yes. Label it as an estimate, include the units, and do not report more digits than you can read from the axis.

What if the caption does not explain an error bar?

Check the Results or Methods section for a definition. Do not assume the bar shows a particular type of uncertainty.