Reading path: Coastal water observation →
The record has stopped speaking
A blank stretch in a coastal time series does not mean the water stayed constant. It means the available record does not show what happened during that interval. A sensor may have failed, a station may have lost power or communications, or data may not have been collected. These are possibilities to investigate, not conclusions to assign without evidence.
Suppose daily temperatures are available on either side of a two-week gap. Drawing a straight line between the surrounding values makes the series look continuous, but it invents a smooth path through conditions that may have changed rapidly. Storms, upwelling, or other events could have occurred during the missing period. The line should not be mistaken for observation.
Look for station notes and parallel records
Check maintenance logs, station notices, and dataset documentation for a reason and timing. Confirm whether the gap applies to the measurement itself or only to a delayed upload. If an independent nearby station or field notebook covers part of the period, it may provide context, but differences in position, depth, and method still matter.
Record the dates and likely cause when known. If the cause is not documented, label it unknown. A visible annotation on a plot can prevent readers from interpreting a missing interval as a quiet phase or a break in the underlying process.
Decide whether estimation serves the question
Interpolation can be useful for a narrowly defined purpose, such as creating a complete input for a model that requires it, provided the approach is justified and the estimated values remain distinguishable from observations. The method should consider the time scale of the process and the length of the gap. A short gap in a slowly varying signal may be treated differently from a long gap across a dynamic period.
Do not use estimated values to claim that an event did or did not occur. If the result depends heavily on the assumed path through the gap, show a sensitivity analysis or leave the interval unfilled. In many descriptive plots, a clean break is clearer and more honest than an unsupported bridge.
Carry the uncertainty into the conclusion
When summarizing a trend, say which portions of the record were observed and whether a gap limits the comparison. A seasonal mean that excludes a missing month may not represent the full season. Report the usable coverage or explain the exclusion so another reader can assess its effect.
A gap is a feature of the evidence, not a nuisance to erase. Investigating its origin and showing it clearly helps separate what the coastal record supports from what remains unknown.
Have a correction or a source to suggest? Send a note →