Attention changes faster than water chemistry. A utility can see a shift in public concern within weeks, but raw water at an intake follows its own schedule, governed by hydrology, upstream land use, and the sampling frequency already in place. When those two clocks are merged in one file, the attention signal starts to read like a measurement. This matters for source-water protection, where boards and customers often ask whether a news item means anything for their supply.
A Circle of Blue piece by Keith Schneider illustrates the gap. Its summary describes national outlets dispatching correspondents to cover "the perilous conduct and unstable condition of industrial agriculture," and frames the coverage as a change in national media interest. The item concerns the volume and direction of reporting. The summary supplies no site-level data, and nothing in it establishes a result for any particular watershed. Read the item as evidence of where questions are heading, and the source note should say so. See Circle of Blue on expanding media interest in agricultural harm.
Three records that answer different questions
Three kinds of record tend to get mixed together. The first is coverage: what outlets report, how often, and which topics draw correspondents. It answers what the public is asking about. The second is regulatory: permits, discharge limits, and the monitoring reports filed against them. It answers which limits apply and whether they were met. The third is operational laboratory data: what was measured, at which point, on which date, by which method, and against which reporting limit. It answers what the water contained at a time and place, within the bounds of that method.
None of the three substitutes for the others. A rise in coverage does not change a permit limit, and a permit does not explain why a board is hearing questions this month. Keeping them apart is a filing decision, not a judgment about whether the coverage is accurate.
Tag each entry before it reaches the board
A useful file tags each entry by record type. A coverage item gets a source name, a link, its publication date, and a note that it contains no measurement. A lab result gets the sample date, the analyte, the method, the reporting limit, and any qualifier codes the lab applied. A permit record gets its effective date and the monitoring frequency it sets. When a board member cites a news item as evidence of a local problem, the staff question is which measured record supports the claim. The answer may be none, and the file should be able to say so plainly.
The date field deserves the same care. A coverage item describes what its author reported at publication, and the conditions it discusses may predate the item by an interval nobody has measured. Record the publication date and do not infer the condition date from it.
Where attention is still useful
Attention is not worthless. It can justify a review of what the utility already holds. Staff can check whether the existing monitoring plan covers the parameters a reader would expect to see, whether the upstream land-use inventory is current, and who holds data for upstream sources the utility does not operate. Those questions can often be answered from records already on hand.
The tradeoff is timing. A sampling program that reacts to coverage may collect data when attention peaks and miss the conditions that produced the signal, while a fixed schedule may miss an event that later reporting describes. A plan that writes down both a fixed schedule and a documented trigger for added sampling keeps the reactive element bounded and reviewable. The trigger should name the condition that would start extra sampling, not the level of public interest in it.
A short template for the file
Each entry can answer five fields in order: what the source says in its own terms, what it does not establish, which monitoring record was checked against it, what that record shows and for which dates, and which question remains open. A coverage entry can usually close only the first two fields. The rest need data that coverage cannot supply, and leaving them blank is more honest than filling them with the strength of the headline.
National reporting will keep moving. The file should record what was measured and what was only reported, and it should keep those two entries visibly separate so the next reader can tell which one a decision rests on.