Ninety-five million salinity records still change meaning when provenance and quality flags are handled differently

A source-aware screen of public estuarine and coastal monitoring found both rising and falling salinity series—and showed that aggregation and flagged observations can change the trend classification.

Wild Lands Ecological Intelligence Editor
Wild Lands Institute · 4 October 2026

Evidence: Peer-reviewed public-data trend screening
Geography: Public estuarine and coastal monitoring networks, 1963–2026
System: Estuaries / coastal waters

WHAT THE EVIDENCE SAYS

The study harmonized about 95 million public salinity observations into 5,383 source-aware station entries from 13 data sources. Of 436 eligible long-term series, 99 showed nominal increases, 76 decreases and 261 no significant trend; false-discovery-rate correction reduced the significant set to 127. Annual aggregation reduced detections, and including flagged observations changed the trend class for 32 of 436 series.

WHY IT MATTERS / OUR READ

Long monitoring archives are not self-interpreting. Provenance, quality-control flags, sampling density and aggregation can change which estuaries appear to be freshening or salinizing. For coastal planning, a trend should travel with its source architecture rather than being treated as a property of the estuary alone.

CAVEATS / UNCERTAINTY

This is a screening analysis of public monitoring records, not an attribution study. Trend direction does not identify the driver, and differences among programs in sampling design and data quality remain part of the inference.

SOURCE

Estuarine, Coastal and Shelf Science — primary study
View original evidence ↗