reflectometry is a diagnostic that probes a plasma like radar: microwaves launched inward travel until they reach the cutoff layer where they can no longer propagate, and the reflected signal carries the electron density profile and its fluctuations.
A wave entering a 📝plasma is slowed by the electrons in its path and stops entirely at a cutoff — a surface where the refractive index falls to zero and the wave turns back. Which surface that is depends on polarization: the ordinary-mode cutoff sits where the wave frequency equals the plasma frequency and therefore depends on electron density alone, while extraordinary-mode cutoffs depend on the magnetic field as well, giving two complementary routes to a given layer. Sweeping the launch frequency walks the reflecting layer through the plasma; the round-trip group delay measured at each frequency locates the layer that returned it, and the set of delays inverts into a density profile. The same phase sensitivity that locates a layer also registers the layer's jitter, so the technique doubles as a turbulence diagnostic — Doppler reflectometry deliberately tilts the beam to backscatter from density fluctuations of a chosen wavelength and returns their velocity.
The distinction against 📝interferometry is locality. An interferometer's beam crosses the whole plasma and returns one line-integrated number; a reflectometer's beam stops at a surface set by its frequency, so each frequency reports a specific radius. Against 📝Thomson scattering, also local, reflectometry trades some absolute accuracy for access: no high-power laser, no in-vessel optics to keep clean, launcher and receiver on the same side of the machine behind a small port, and a sampling rate fast enough to feed real-time control. That is a reactor argument, and it is why 📝ASDEX Upgrade, Tore Supra, 📝DIII-D, 📝JET and NSTX have all fielded reflectometers for routine density profiles and why 📝ITER carries three separate reflectometry systems in its design. 📝Jon Hillesheim, now at 📝Commonwealth Fusion Systems, built one such instrument earlier in his career: a multichannel, frequency-modulated, tunable Doppler backscattering and reflectometry system published in 2009.
📝SPARC's version is designed but unbuilt. Edge scanning reflectometry for density profile measurement on the SPARC tokamak, led by 📝Yijun Lin in 2024, sets out a frequency-modulated continuous-wave system sweeping 18 to 90 GHz across four bands, using both ordinary mode and the left-hand extraordinary-mode cutoff to cover densities from roughly 4 × 10¹⁸ to 4 × 10²⁰ per cubic metre at 12 📝tesla — from the far 📝scrape-off layer to the top of a projected 📝H-mode 📝pedestal, with the data feeding plasma control. The coverage figures are design specifications supported by electromagnetic modeling, not measurements.
Edge density is something SPARC has to steer on, not merely record afterward, which is why the reflectometer is being designed to deliver a profile in real time. The 18–90 GHz system exists on paper and in simulation; nothing has yet reflected off a SPARC plasma.