I thought I had asked this question before, but I can't find it now so I will ask here.
I have asked a related question in Physics SE: Do III-V based photovoltaics “glow” (photo-luminesce) when illuminated but not loaded? and the answer (as expected) is yes.
New Question: Have spacecraft ever been detected by direct bandgap photoemission from III-V solar panels?
The question When there is more power than needed, are satellite photovoltaics shunted or open circuit? examines what happens to all of the energy deposited in photovoltaic cells by sunlight that has gone into the production of electron-hole pairs, which is roughly a third of the incident energy.
If you leave the PV cell disconnected, recombination will liberate that energy as heat, and the cells can get even hotter if not cooled properly. However, connect the PV cell to a shunt resistor and the heat will be deposited in the resistor which in many cases is more convenient and easier to cool.
However if the cells are made from high quality direct bandgap III-V material rather than silicon, a significant fraction of the recombination events can produce light at the bangap energy, which can be in the near IR or even visible for some materials.
Panels with III-V junctions, illuminated by Sunlight at 1 AU could be emitting of order 100 W/m^2 of light in a narrow band of wavelengths, something that would stand out like a sore thumb with the proper filter.
If they don't use shunting, you might even be able to tell if the spacecraft is loading their panels or leaving them open in times of low power usage.
Has this ever been reported or observed? Ever talked about?