A tide gauge, an offshore buoy, a river sensor, and a water-quality sampling station don’t agree on much: not format, not schedule, not units, not how they flag a bad reading. Before any of it can become a useful picture of a bay, it has to be put on common footing. This post walks through how we bring these sources together and hold every one of them to the same quality standards, so a number on our weekly read means the same thing no matter where it came from.
Where the data comes from
As of June 2026, DataShoal draws on two dozen public data feeds across five kinds of sources for Humboldt, Tomales, and Morro bays: offshore buoys run by NOAA’s National Data Buoy Center, NOAA tide gauges, USGS river gauges, continuous water-quality sensors operated by universities through the regional CeNCOOS network, and weekly harmful-algae sampling by the state’s CalHABMAP partners at UC Davis’s Bodega Marine Laboratory, Cal Poly Humboldt, and Cal Poly San Luis Obispo. Together that’s around nine million measurements and more than a decade of history, refreshed every week.
None of this would exist without the agencies, universities, and volunteers who keep those stations running and publish the readings openly.
Why “common footing” is most of the work
Each source arrives in its own format, on its own schedule, in its own units, with its own way of saying “trust this number” or “don’t.” A buoy reports hourly; a sampling station reports weekly; a tide gauge reports every six minutes. One feed marks a suspect reading with a code, another leaves a gap, a third quietly reports a value that is physically impossible.
If you put those side by side without doing anything, you get a picture that looks complete and isn’t. So every reading that enters the project is converted to shared units, stamped with when it was observed and when it arrived, and kept alongside a record of exactly where it came from, so any number can be traced back to its source.
One quality standard for everything
Every reading, from every source, is checked against the same quality standards oceanographers use. The tests come from QARTOD, the quality-control manuals published by NOAA’s ocean-observing program, and we run them with the open-source software library the ocean-observing community maintains for exactly this purpose. The tests ask ordinary, skeptical questions. Is this value physically possible? Did the sensor jump implausibly fast? Has it flatlined?
Readings that fail don’t disappear. They’re flagged and kept, so the record shows what was measured and what was doubted. When the weekly read says conditions were normal, that claim stands on data that has been screened the same way whether it came from a federal buoy or a volunteer’s weekly sample.
That’s the standard we hold ourselves to as well: every number in a weekly report has to trace back to verified data, and where confidence is lower, the report says so.