Reading path: Underwater images and evidence →
Find the data behind the image
A marine visualization can make patterns easy to notice, but the graphic’s appearance does not establish that its message is sound. Start by locating the source: a paper, monitoring program, repository, or model description. Check the date range and geographic coverage. A map may show one survey, a long record, or an estimate across unsampled water, and those cases call for different interpretations.
Read the caption and accompanying text. They may explain who collected the data, how observations were grouped, and whether the figure is measured, modeled, or both. If a graphic is separated from its original explanation, treat it as a lead. Search for the underlying source before repeating a claim based on its colors or headline.
Check units and scale
Look at axis labels, color bars, and units. A plotted temperature, concentration, count, or rate means little if the unit is absent or unfamiliar. On a map, check whether the color scale is continuous or divided into categories. Ask where the scale begins and ends: a narrow range can make a modest change look dramatic, while a broad range may hide local variation.
Inspect the time and space scales too. Does the horizontal axis cover days, seasons, or decades? Are values averaged by month or shown at individual observations? Does a map use a projection that distorts area, and are sampling stations dense enough to support the apparent surface pattern? These questions help you separate a display choice from an environmental pattern.
Look for gaps and transformations
Find out how missing values are represented. Blank regions, zero values, and interpolated estimates are not interchangeable. A gap could mean no sample was taken, a sensor failed, or data were excluded after quality checks. If a map fills gaps with estimated values, the method and uncertainty should be available somewhere in the source material.
Check for transformations such as smoothing, normalization, or averaging. These operations can reveal broad patterns but may conceal brief events or local extremes. For example, a monthly average can dampen a short temperature spike. Ask whether the visualization answers the same question as the headline. A graphic showing association across sites does not, on its own, establish the cause of the pattern.
Match the graphic to the claim
State what the figure directly shows before interpreting it. “The plotted values rise across the sampled dates” is a description; “warming caused the population change” is a causal explanation that needs more evidence. Check whether the underlying analysis tests that explanation or whether the graphic is offered as context. Keep uncertainty and sample boundaries visible in your summary.
A useful graphic-reading checklist is source, units, scale, time and location, missing data, transformations, and claim. Record any unresolved issue instead of filling it with an assumption. If you can trace the data and understand how the image was built, the visualization becomes a better guide to the evidence. If not, return to the paper or dataset before drawing a conclusion.
Have a correction or a source to suggest? Send a note →