EFDA-JET-PR(10)05

Investigation of the TS/ECE Temperature Measurement Discrepancy at JET Using the Core LIDAR TS Diagnostic

The present work is motivated by a long standing discrepancy between the electron temperature measurements of Thomson Scattering (TS) and Electron Cyclotron Emission (ECE) diagnostics for plasmas with strong auxiliary heating observed at both JET and TFTR, where in some cases the TS electron temperature measurements can be 15-20% lower than ECE measurements. This problem is a significant concern for the fusion community as highlighted in [1]. Recent analysis based on ECE results at JET have suggested that the assumption of a Maxwellian electron velocity distribution may not be valid under the influence of high levels of NBI and ICRF heating. Such results have indicated distortions to the low-energy part of the distribution, rather than simply the generation of a tail of fast electrons. A non-Maxwellian distribution function was proposed by Krivenski, which appeared to resolve the difference between the ECE and TS central temperatures but a physical mechanism responsible for such behaviour was not identified. In this paper, the JET core LIDAR TS measurements have been utilised in an attempt to detect the presence of non-Maxwellian distributions. As part of this work, a model was developed to evaluate the theoretical number of scattered photoelectrons in each spectral bin of the core LIDAR TS system for an arbitrary electron distribution. Using this model, the experimental measurements of the LIDAR system, under numerous heating conditions, have been compared with theoretical data simulated from the non-Maxwellian bulk distribution and a 'high energy tailed' Lorentzian distribution. In this way, efforts were made to isolate the most likely velocity distribution for certain heating conditions and determine if any detected deviations from a Maxwellian could be related to the TS/ECE discrepancy. Additional analysis was carried out to identify possible calibration inaccuracies with the LIDAR system and determine if their effect could contribute to the TS/ECE discrepancy.
Name Size  
EFDP10005 1.59 Mb