Quantifying noise sources in the KSTAR 2014 Thomson Scattering system from the measured variation on electron temperature

Abstract

With the Thomson scattering (TS) system in KSTAR, temporal evolution of electron temperature (T e ) is estimated using a weighted look-up table method with fast sampling (1.25 or 2.5 GS/s) digitizers during the 2014 KSTAR campaign. Background noise level is used as a weighting parameter without considering the photon noise due to the absence of information on absolute photon counts detected by the TS system. Estimated electron temperature during a relatively quiescent discharge are scattered, i.e., 15% variation on T e with respect to its mean value. We find that this 15% variation on T e cannot be explained solely by the background noise level which leads us to include photon noise effects in our analysis. Using synthetic data, we have estimated the required photon noise level consistent with the observation and determined the dominant noise source in KSTAR TS system.

Estimation of reliable range of electron temperature measurements with sets of given optical bandpass filters for KSTAR Thomson scattering system based on synthetic Thomson data

“…and the number of photo-electrons N s from an APD detector can be written as Integrating photo-electrons with respect to the scattered wavelength λ s and multiplying it with the APD gain factor G provides the electronic signal that we can read using a digitizer, i.e., equation (2.1). For the purpose of this paper, actual value of G is irrelevant because the KSTAR TS system is dominated by the photon noise [6] which means that noise levels are determined by the number of photo-electrons. We note that our forward model can act as a likelihood function in the frame of Bayesian probability theory [10].…”

“…Then, results are compared with the prescribed profiles of electron density and temperature to estimate the reliability of the KSTAR TS system. synthetic TS data [6] with prescribed profiles of electron density and temperature. The forward model is based on the TS system parameters used during the 2016 KSTAR campaign.…”

“…Treating the synthetic TS data as if they were experimentally measured ones, we can obtain temperature and density of electrons by using a look-up table method [6,7] which, then, are compared with the prescribed temperature and density. Figure 2 schematically describes such a procedure.…”

“…Such fractional errors provide quantitative magnitude of random errors but do not present bias errors. With the aim of obtaining both bias (if finite) and random errors, we have developed a forward model of the KSTAR TS system and generated synthetic TS data [6] with prescribed profiles of electron density and temperature. The forward model is based on the TS system parameters used during the 2016 KSTAR campaign.…”

Feasibility study of direct spectra measurements for Thomson scattered signals for KSTAR fusion-grade plasmas

“…Based on the KSTAR TS system [7] used during the 2016 Campaign [15], the expected number of photons [16,17] at a certain scattered wavelength λ denoted as µ λ can be expressed as…”