Blue-green algae vs. golden algae: reading the signal in satellite imagery

Why this question comes up every August
Someone on plant staff pulls up a true-color satellite scene of the reservoir, sees a greenish-brown streak off the north arm, and asks the obvious question: is that the toxin-producing stuff, or is it golden algae again like two summers ago? True color alone won't answer it. A chlorophyll bloom is a chlorophyll bloom in RGB, whether it's Microcystis or a chrysophyte. The difference lives in bands most people never look at.
The pigment that gives cyanobacteria away
Blue-green algae are cyanobacteria, and they carry a pigment called phycocyanin that chrysophytes like golden algae lack. Phycocyanin absorbs light in a narrow band around 620 nm, right between the chlorophyll reflectance peak in the green and the red-edge bump near 700 nm. Sensors tuned to catch that dip can flag cyanobacteria specifically, instead of just flagging "something photosynthetic is out there." Golden algae (commonly Prymnesium parvum in the reservoirs where it causes problems) runs on chlorophyll-a and chlorophyll-c plus accessory pigments closer to the diatom and dinoflagellate family, with no phycocyanin signature and no 620 nm absorption dip. A bloom that shows strong chlorophyll reflectance but no phycocyanin signal is pointing you toward golden algae or another non-cyanobacterial group.
Surface behavior is the second clue
Cyanobacteria carry gas vesicles that let cells regulate buoyancy, which is why cyanobacterial blooms so often show up as a thin, streaky, windrow-pattern scum sitting right on the surface, especially after a few calm days let cells float up and wind pushes them into a line along the shore or the intake crib. Golden algae tends to stay mixed through the water column and rarely forms that same paint-on-the-surface look, even during a heavy bloom. If imagery shows a tight, bright, wind-aligned streak, lean cyanobacteria. If it shows a broader, more diffuse discoloration without the streaking, that's more consistent with golden algae or another flagellate bloom. It's a tendency, not a rule. Wind, mixing, and bloom age all change how either group presents on a given pass.
Where imagery stops and the lab starts
None of this replaces a grab sample. Spectral signature and surface pattern narrow the field and tell you where to send a bottle and how urgently, but the only way to confirm species, genus, or toxin presence is microscopy, qPCR, or a pigment assay run by your lab or state program. What satellite imagery is good for is catching the bloom forming and telling you where to look before it's close enough to the intake that you're reacting instead of planning. That's the gap a lot of source water programs have: no early warning that something's forming three miles up the reservoir until it's already at the crib.
Triage: how to read today's pass
If you're trying to sort out what you're seeing on a given pass, treat it like this: chlorophyll signal plus a phycocyanin dip plus a surface streak pattern says send a sample and start watching your intake depth. Chlorophyll signal without the phycocyanin dip, in a more diffuse pattern, is probably something other than the cyanotoxin-producing group, though still worth a sample if you've had golden algae history or a fish kill risk at that reservoir. Either way, the imagery's job is to tell you something changed and roughly where, early enough to adjust intake timing or get ahead of a public notice conversation instead of finding out when the taste-and-odor complaints start coming in.
That's the gap Algal Bloom Warning is built to close: a daily-in-season read of your reservoir's spectral signature that flags bloom extent before it drifts toward the intake, with a trend note so you're not starting from zero every time someone asks what that streak on the imagery is.
If you're tired of guessing at bloom ID from a true-color screenshot, see how a daily pass over your own reservoir changes that conversation.