Could radiation pressure noise become limiting again if squeezed light's momentum uncertainty grows at higher frequencies?

Your question cuts to something subtle: as we use squeezed light to reduce quantum noise at low frequencies in gravitational-wave detectors, does the momentum uncertainty of that squeezed state grow at higher frequencies in a way that reintroduces radiation pressure as a limiting factor?

The short answer is: not quite as a fundamental problem, but frequency-dependent squeezing is essential to avoid this exact trade-off.

Why the trade-off exists

The Heisenberg uncertainty principle is inescapable: if you squeeze position uncertainty below the standard quantum limit (SQL) at one frequency, you *must* anti-squeeze momentum uncertainty [AG-2024.04-408]. The detector becomes more sensitive to weak gravitational waves (which live at specific frequencies like 100 Hz for LIGO) but more vulnerable to radiation pressure noise—the recoil of photons pushing the test mirrors.

In older broadband-squeezed approaches, this was indeed a problem: suppress quantum noise everywhere equally, and you worsen radiation pressure at all frequencies.

How frequency-dependent squeezing solves it

The modern solution is to rotate the squeezing ellipse with frequency, so that:

• At low frequencies (100–300 Hz), you squeeze *position* (reducing shot noise from photon counting)
• At high frequencies, you squeeze *momentum* (reducing radiation pressure)

This requires filter cavities—narrow-band optical cavities that delay and rotate the squeezed light's quantum state as a function of signal frequency [AG-2024.01-1430, AG-2024.04-2117]. LIGO's A+ upgrade demonstrated this working in practice, achieving broadband sensitivity improvement *while staying below the SQL* [AG-2024.04-408].

The practical constraint

The trade-off doesn't disappear; it's architected away. But it does impose engineering costs: Einstein Telescope's design currently requires *two filter cavities* to handle its dual-recycled geometry [AG-2024.04-2117]. Novel schemes like EPR squeezing (entangled-pair squeezing) may eventually eliminate one filter cavity, but this remains an active research frontier.

The deeper insight from [AG-2024.08-416] is that once you inject squeezed states, accounting for all loss channels, feedback control, and classical noises becomes paramount—the quantum state alone doesn't guarantee sensitivity; its preservation through the interferometer does.

So radiation pressure doesn't "become limiting again" in the sense of a reappearing floor—it's always there as a quantum trade-off—but frequency-dependent squeezing ensures you don't sacrifice one band for another.