Sampling
Where to sample first
Rank the sites most likely to be a problem this week, so limited crews cover the water that matters most.
Water warning lab · Monterey Bay
Know which water
to check first.
We turn the monitoring data you already collect into a clear, ranked answer — which beaches, rivers, and lakes to sample before they become a problem.
You can’t test every beach, lake, and river. We tell you where to look first — and we prove it beats the rule you use now.
Early warning
Warnings you can trust
Flag likely blooms, closures, and contamination before they happen — with an honest read on how sure we are.
Signal checks
Is the signal real?
Test whether a new measurement actually predicts trouble, or just looks like it does.
Proof, in plain numbers
~⅓ more
Contamination caught in California — the same number of samples, just aimed better.
8 in 10
New Florida red-tide outbreaks flagged while checking only one site in five.
no local data
A California-trained model ranked shellfish-toxin risk on the coast of Ireland — and beat every published method.
The full record — including what didn’t work — lives in the evidence library.
“We publish the results that didn’t work, too.”
Knowing what to stop doing is part of what a pilot buys you.
How we work
01 — Baseline
It has to beat the simple rule
If “it rained, so post a warning” does just as well, we say so. We don’t dress up the obvious as AI.
02 — Red-team
Someone tries to break it
Every result gets a second, skeptical review before we stand behind it. Weak claims get narrowed or dropped.
03 — On the record
We show the misses
The dead ends go on the record, not just the wins. That’s how you know the wins are real.
Same method, many kinds of water.
- Beaches & estuaries
- Rivers & streams
- Lakes & reservoirs
- Shellfish & harmful algae
- Low-oxygen zones
- Toxicity & wildlife
Start a pilot
Four weeks to a clear yes or no.
Pick one water system and one decision. We’ll tell you whether your data can support it — and what the smallest useful tool would look like.
- 1Pick one decisionWhere to sample, what to watch, or a signal to check.
- 2Share your dataWhatever you already collect — we’ll tell you what’s usable.
- 3We test it fairlyAgainst the simple approach you’d use anyway.
- 4You decide what’s nextA tool, a report, or an honest “not yet.”
Tell us about your water.
One water system, one decision, and roughly what data you hold. We’ll reply with whether it can work — and the smallest first step.