Earth-system data is now widely available — through NASA’s Earthdata programme and many other sources — and widely used in public communication. Maps, charts and visualisations appear in news reports, advocacy materials and policy documents. That visibility is valuable and risky, because data does not speak for itself. Scale, uncertainty, colour choices and the difference between a model and a measurement all shape what a reader takes away, and all are easy to get wrong.
Data is not the same as measurement
Much of what circulates as Earth observation is derived from models or remote sensing rather than direct measurement. A temperature map may combine satellite retrievals, ground stations and reanalysis; a flood-risk layer may be the output of a hydrological model. These are legitimate and useful, but they carry assumptions and uncertainty that a clean graphic conceals. Presenting modelled output as if it were a direct observation misleads, even unintentionally.
The problem of scale
Earth-system data is often produced at coarse resolution and then displayed at finer resolution than it supports. A global dataset with cells tens of kilometres wide, rendered on a city map, implies precision that does not exist. Readers may conclude that their neighbourhood has a particular risk when the underlying data cannot distinguish one neighbourhood from another. The honest approach states the resolution and avoids visual over-claiming.
Uncertainty and how to show it
Every estimate has uncertainty, and most public graphics hide it. A single line on a chart suggests more confidence than an ensemble of model runs supports. Showing ranges, error bars or scenario spreads is less satisfying visually and more truthful. The gap between scientific practice, where uncertainty is central, and public communication, where it is often stripped away, is a major source of misunderstanding.
Colour, projection and framing
Design choices carry meaning. A colour scale can make a modest change look alarming or a large one look benign. A map projection can distort areas, making some regions appear larger than they are. Cropping a chart can exaggerate a trend by truncating the vertical axis. These are not trivial details; they are where a truthful dataset becomes a misleading message. Good practice uses perceptually uniform colour scales, honest axes and stated projections.
Attribution and causation
Seeing two variables move together is not the same as knowing that one causes the other. A visualisation that overlays emissions and temperature invites a causal reading that the data alone may not support. Attribution science has become more rigorous, but it works through specific methods and stated confidence, not through the visual juxtaposition of two trends. Readers should look for explicit causal claims and their basis, not infer them from a graphic.
Who made the map, and why
Data visualisations are produced by someone, for a purpose. That is a reason to read them as arguments as well as evidence. Advocacy groups, agencies and media outlets select what to show and how. Checking the source, the method and the underlying data is the difference between being informed and being persuaded. Where the data and method are open, that check is possible; where they are not, caution is warranted.
The value of open data and open methods
Openness is the main defence against misleading visuals. When data and code are available, others can reproduce a figure, test a different scale, or check an assumption. This is why the open-data movement matters for public understanding, not only for research. A claim that can be checked is more trustworthy than one that cannot, regardless of who makes it.
How to read an Earth-system graphic
Ask a short set of questions. Is this a measurement or a model? At what resolution was it produced, and is the display honest about that? Is uncertainty shown? What is the colour scale, and is the axis truncated? Who made it, and can the underlying data be checked? These questions take a minute and prevent most misreadings. They also make the reader harder to mislead, which is the point.
Reading the data with care
Earth-system data can inform public debate, and it can distort it. The difference lies in how carefully it is presented and how critically it is read. Openness and transparency are what allow that critical reading to happen, and they deserve as much attention as the data itself.
The role of metadata
Behind every usable dataset is metadata: what was measured, when, by what instrument, at what accuracy and under what licence. Metadata is what allows a user to decide whether a dataset fits their question, and it is frequently incomplete. A file named “flood risk” tells a reader almost nothing. This is why open-data advocates treat documentation as part of the data, not a courtesy. Where metadata is missing, users guess, and guesses propagate into decisions. Investing in documentation is unglamorous, and it is the difference between data that can be reused and data that can only be displayed.
Sources
- NASA. Earthdata. earthdata.nasa.gov
- IPCC. Assessment Reports — on uncertainty and attribution. ipcc.ch
Image: Wendy Cope · CC BY 2.0 · via Wikimedia Commons.