Essays from the case file.

The claim has three qualities that make it unusually sticky online: it is instantly legible, absurd enough to share as a joke, and structurally similar to serious conspiracy narratives. You do not need specialist knowledge to understand the premise. That makes it easy to compress into a meme, a short video or a single sentence.

The problem begins when the joke’s self-protecting logic is taken seriously. If photographs are fake, travellers are paid, residents are actors and every institution is compromised, then almost any counterexample can be absorbed into the theory. This creates the appearance of explanatory power at the cost of testability.

That is why the best response is not to produce an even larger pile of screenshots. It is to ask what would distinguish the theory from its alternative. A good explanation should make risky predictions. If no imaginable evidence could count against a claim, its ability to “explain” observations tells us very little.

A satellite image is evidence, but not magic. A single JPEG on a website can be copied, edited or mislabelled. The stronger evidence is the observation system behind it: repeat passes, known instruments, metadata, multiple agencies, weather patterns, ground observations and datasets that can be compared over time.

This distinction matters because healthy skepticism should be aimed at provenance, not aesthetics. “It looks fake” is a weak test. “Where did this image come from, what sensor produced it, when was it acquired, and does independent data agree?” is a much stronger one.

NASA’s 2026 imagery of Cyclone Narelle illustrates the approach. The image is tied to VIIRS on NOAA-21, a date, an observed storm track and ground impacts. It can be compared against Australian meteorological observations and other records. Convergence is more informative than the visual alone.

Psychological research has long examined the “illusory truth” effect: repetition can increase familiarity, and familiarity can influence truth judgments. That does not mean repetition automatically persuades everyone, or that the effect is equally large in every context.

The practical implication is simple: corrections should not spend most of their space dramatising the false claim. State the accurate information clearly, identify the claim briefly, explain the evidence, and finish with the correct model. This website therefore places the reality check at the top of every page.

Search is not the same as research. If a person starts with a conclusion and only asks questions that assume the conclusion, modern search systems can provide an endless supply of confirming material. Proper inquiry needs an adversarial step: identify the strongest evidence against your preferred interpretation.

That means tracing claims back to primary sources, reading methods rather than only headlines, checking later replications, and distinguishing a result that is statistically associated with an outcome from an intervention that causally changes it. The 2014 analytic-thinking study and its later registered replication are a useful example of why a source trail should include updates, not only famous findings.