The quality of Thomson scattering (TS) profiles relies on high-quality calibrations since each spatial point is essentially an independent measurement. For the spectral calibration of TS diagnostics, several techniques exist. Typically, these include a known light source and involve changes to the optical setup to observe that light source. Two assumptions are frequently made: The spectral response of the diagnostic is stable in time and the required changes to the optical setup only lead to negligible errors in the calibration. The temporal stability is questionable for larger fusion experiments, where the harsh environment leads to coating and degradation of optical components. Both the stability and validity of the spectral calibration can be checked by in-situ calibrations. Here, the calibration is performed either during operation or in between experiments without any modifications to the diagnostic. So far, two promising candidates have been proposed: dual-wavelength TS and Rayleigh scattering on, for example, argon gas using a tunable optical parametric oscillator. In this contribution, we introduce a new method for an in-situ spectral calibration using experimental data from plasma measurements. This method was developed for polychromator-based TS diagnostics, but can probably also be adapted to spectrometer-based systems. A forward model is used to predict the signals measured by the diagnostic (assuming a certain spectral calibration) and these predictions are then compared with experimental data the diagnostic has acquired. Using an optimizer, the spectral calibration is varied until predicted and measured data agree. Rough knowledge of the polychromator design is sufficient to find an estimate for the calibration of the diagnostic. We demonstrate, using polychromators from Wendelstein 7-X as example, that even data measured in the past can be calibrated with this technique.
Dual-laser calibration of Thomson scattering systems in ITER and RFX-mod