Reconstruction of the emissivity and flow for Doppler coherence imaging spectroscopy (CIS) on J-TEXT

•Doppler coherence imaging spectroscopy (CIS), as novel diagnostics, has been applied several plasma devices to provide the information about the emissivity and flow of impurity ions, such as DIII-D, MAST, HL-2A, J-TEXT, etc. However, the information extracted from the diagnostic is line-integrated, so it is necessary to infer the local emissivity by using the inversion algorithm. Especially the experiments on J-TEXT, the aim is to study the effect of the movement of strike point on transport. Quantitative description of the movement of strike point is very important for experimental transport. At present, the inversion results have been used in experimental analysis and simple conclusions have been obtained. “H. Wang, Y.F Liang, S. Xu, Z.H. Jiang, Y.H. Feng, K. Alexander, et al. Influence of the X-point location on the edge plasma transport in the J-TEXT tokamak with a high-field-side single-null divertor, Plasma Sci. Tech. (2021).” According to the diagnostic characteristics of CIS, we choose two inversion algorithms, Tikhonov method and minimum fisher information (MFI) method, and test the algorithm be two phantom models. And the reconstructed results of emissivity and flow was provided, which could quantitatively express the movement of the strike point, and has important reference value for the study of plasma rotation. Doppler coherence imaging spectroscopy (CIS) has been installed on J-TEXT tokamak to provide a wealth of information for carbon impurity ions (CIII). To obtain the local emissivity and flow from the line-integrated information measured by the camera, the tomographic inversion needs to propose. However, besides the ill-conditioned nature of the inversion, the distribution of CIII (only in the narrow scrape-off-layer (SOL) and around the divertor) makes the inversion more challenging. And the error in the emissivity inversion may propagate to the flow inversion. In this article, we have introduced the two classical tomographic algorithms, Tikhonov regularization, and minimum Fisher information regularization. To estimate the uncertainty of these algorithms and study the error sources, two typical CIII distribution models were chosen to make benchmarks. Finally, these methods were applied to a J-TEXT experiment, demonstrating the emissivity and flow distribution in SOL and around the divertor.