Where satellites are seeing cyanobacteria.

This map cannot answer that, and it is the most important thing to understand about it. A satellite reads only the surface of the water; wind mixes a bloom down through a lake, so a bloom that is genuinely present can be missed entirely and the lake will read clear here. Readings below roughly 10,000 estimated cells/mL are beneath the sensor’s detection limit — that is “we could not see anything”, not “there is nothing there”. Your state’s cyanobacteria programme issues the advisories, samples the water, and is the authority on whether a lake is safe to use. This map is a pre-screening indicator that says where to look.

No, despite “blue-green algae” being the common name. Cyanobacteria are bacteria. The distinction matters because the toxins that make a bloom dangerous — microcystin among them — are produced by the bacteria, and guidance written about algae does not necessarily apply. Plenty of green water is harmless algae, and plenty of cyanobacteria blooms are themselves harmless. Colour is not a reliable guide either way.

No. The satellite measures a pigment that cyanobacteria contain, not a toxin. The index cannot distinguish a toxic bloom from a harmless one and says nothing whatever about microcystin concentration. A high reading means there is a lot of cyanobacteria at the surface — a good reason to check your state advisory and to stay out of visible scum, and not a determination that the water is poisonous. Only sampling the water answers the toxicity question.

They are estimates converted from a pigment index using a published relationship (Lunetta et al. 2015), not counted cells. That conversion carries roughly 29% mean absolute percentage error, so a figure of 100,000 cells/mL should be read as a number with real spread around it rather than as a measurement. The two density landmarks the bands use — about 20,000 and about 100,000 estimated cells/mL — are the values US state programmes most commonly cite, and they vary by state. Check your own state’s thresholds; those are the ones that apply where you are.

Most likely it is too small. The sensor resolves at 300 m, and a lake needs several clear water pixels away from its own shoreline to be read at all — roughly 900 m across. Around 2,300 U.S. lakes across 46 states meet that bar. Rivers, ponds and most reservoirs are absent entirely, and their absence says nothing whatever about their condition. A lake can also be present but unassessable in a given week, in which case it is drawn with the reason rather than left off.

That we have no reading, and the map says which of six things prevented one: ice or snow on the lake, cloud over it, the lake reading as mostly land in the satellite product because it is drawn down or dry, a lake too small to resolve, a lake outside the product’s footprint, or our own failure to retrieve the imagery. That last one is ours rather than weather, and it is labelled separately for exactly that reason. An unassessable lake is never evidence of a clear lake — those are different facts, and this map keeps them apart deliberately.

The map renders NASA and EPA’s 7-day composite, which is published weekly and lags a few days behind the end of its own window. Our sampler runs daily so a late composite is picked up promptly, but the underlying observation does not move faster than the satellite product does. The window the map covers is shown on the page. NASA publishes the underlying product as preliminary and for evaluation purposes only.

The impaired waters map shows what states have formally told EPA about their own waters under the Clean Water Act — a regulatory judgement made over assessment cycles that run for years, and one of its listed causes is nutrient and algal growth. This map is a satellite reading of a single recent week. A lake can be listed as impaired for nutrients and read clear here, or read high here and never have been assessed at all. Neither supersedes the other, and neither is an advisory.

Not yet, and we say so rather than borrowing someone else’s numbers. Published agreement figures for satellite cyanobacteria screening come from the research literature and describe other lakes in other seasons — they are not this map’s measured performance. A one-season retrospective comparing what this map published against state advisory records is planned; until it runs, treat any agreement statistic you see quoted as being about the technique in general, not about this map.

Yes. Satellite cyanobacteria screening is one of more than a dozen federal inputs Driftwise watches on the waters a customer is responsible for, alongside USGS gauge readings, the National Water Model, drought classification, NOAA flood forecasts, wildfire proximity and EPA water quality designations. In the product a bloom reading is resolved to the specific access points and reaches a crew manages, so it becomes a factor in what a steward is told to do this week rather than a page somebody has to remember to check.