Validation of delay dispersion estimation for coherence imaging spectroscopy

Coherence Imaging Spectroscopy (CIS) is a camera-based polarization interferometry technique that provides high-spatial-resolution 2D measurements of spectroscopic quantities. Its most common application is in the field of fusion plasma diagnostics. Like other systems based on birefringent crystals, the CIS diagnostic has a characteristic delay dispersion or group delay, i.e., how much the interferometer phase response varies with small changes in wavelength. This dispersion is one of the key system parameters that must be precisely characterized in order to derive physics-relevant quantities (such as ion flow velocity and temperature) from the raw diagnostic signal. This paper reports on the validation of delay dispersion estimation methods, based on extrapolations and interpolations, for wavelength ranges hard to access with conventional calibration sources, such as spectral lamps, diode lasers, or monochromators. The methods under investigation use either the simulation of the system response adopting simple models or power-law fits of the available delay dispersion measurements. The tests are performed using two continuous-wave tunable lasers that, together, cover the range 450-750 nm without any gaps. Moreover, the analysis is repeated for three different CIS systems featuring different crystals, imaging lenses, and cameras, revealing that the crystal alignment and lens quality can substantially influence the precision of the estimation. The smallest deviation (<2%) between the estimated and measured delay dispersion is obtained with the simplest CIS setup, less prone to hardware imperfections. The systems featuring a more complex setup show deviations that can reach 20%, including spatial structures difficult to capture with the tested estimation methods.