Statistical method for efficient determination of electron temperature from pulse height analysis of soft x rays in Heliotron E.

A pulse-height-analysis (PHA) technique of soft x rays is applied to Heliotron E experiments. The primary interest of this measurement is the determination of the electron temperature. Since the number of photons measured in a PHA system is extremely restricted owing to a finite processing rate of the system, it is most important to use the full amount of information involved in the spectral data. The statistical estimator from the maximum-likelihood method is efficiently used for this purpose. The estimated temperature is in good coincidence with the temperature by the electron-cyclotron-emission measurement, which is calibrated by the laser Thomson scattering. A check of reliability of the estimated electron temperature is developed from a statistical test of goodness of fit. The identification of impurity lines buried in a thermal spectrum is also discussed in the spectra accumulated through several or several tens of plasma discharges. The line emissions from Si, Cl, Ca, and Ti are identified. A systematic error in the estimated temperature due to the impurity lines is evaluated. The contributions from Cl and Ti cannot be neglected in the temperature estimation from a spectrum accumulated through several currentless ECH plasmas in Heliotron E. A removal of those contributions to the temperature estimation is successfully demonstrated. The electron temperature with several percent uncertainty is obtained by PHA every 10 ms during a Heliotron E discharge. [ABSTRACT FROM AUTHOR]