Follow change through time

The documented AHI baseline uses ten-minute full-disc scans and 2.5-minute Japan and target-area scans, with housekeeping exceptions. Faster regional scans trade the geographic scope of that observation for more frequent updates. An instrument schedule describes intended sampling; a delivered dataset must still be checked for gaps.

For an ideal uninterrupted hour, these intervals correspond to six full-disc scan cycles or twenty-four regional cycles. These are rates calculated from the baseline, not a promise that every hour contains those exact delivered products. Counting both endpoints of a chosen interval can also change the number of timestamps without changing the scan rate.

In an original teaching example, a recognizable cloud pattern shifts 12 kilometers between observations taken ten minutes apart. Dividing distance by elapsed time gives an average apparent displacement rate of 1.2 kilometers per minute, or 72 kilometers per hour. These fictional numbers are not a measurement of the cyclone pictured elsewhere in this course.

Now suppose the middle observation in a sequence is missing, so the two available frames are twenty minutes apart. Using ten minutes in the denominator would double the inferred rate. Read the actual timestamps and their time zone before calculating. A smooth animation can conceal irregular intervals if every frame is displayed for the same duration.

Use the actual observation interval

Cloud shapes can grow, fade or reorganize between frames. Following a bright feature does not necessarily follow the same parcel of air, and a pattern’s displacement does not directly establish surface wind speed. Feature tracking is useful when its assumptions and limits remain visible.

If a gap spans rapid development, keep it marked as missing rather than inserting an invented observation. A forecast or interpolation can be useful, but label it as such. A weather record becomes more informative when measured frames, estimates and missing intervals remain distinguishable.