Compare only valid observations of normally lit locations when assessing new darkening. Distinguish image-supported change from a field-confirmed electricity outage.
A known horizontal mirror reverses landmark order and directional relationships. Account for that transformation before using a panorama to select a location.
A terrain grid supports measurements at its stated scale and uncertainty. It does not automatically establish human-scale passage width, accessibility or the identity of a hidden structure.
An unfamiliar shape is an observation that needs testing, not a completed identification. Compare independent viewpoints, time evolution and measurement uncertainty before describing a rigid object.
A sampled signal can appear at a different frequency from its source. Compare candidate frequencies across the supplied sample rates rather than identifying a source from one apparent rhythm.
Sensitivity and the fraction of alerts that are true answer different questions. To interpret an alert, combine the supplied target count, missed targets and false alerts instead of repeating an accuracy slogan.
First identify what an image band measures and how it is displayed. False colour in a reflected-light product does not become calibrated heat evidence because the palette includes red.
A cluster of reports is not automatically a synchronized event. Normalize clocks, apply the predeclared timing rule and examine how the reports were selected before discussing a pattern.
A before-and-after comparison needs compatible dates, measurements and observation masks. Calculate change within the area valid in both images rather than using an attractive full-frame denominator.
An apparent receive-before-send contradiction may come from calendar or timezone conventions. Preserve the original notation and apply each declared conversion separately before assessing the sequence.