Characterizing plasma with emission tomography-Feasibility study on synthetic and experimental data.

We present a feasibility study on different tomographic algorithms to overcome the issues of finite sets of projection data, limited viewing angles, and noisy data, which cause the tomographic reconstruction to be an ill-posed inversion problem. We investigated three approaches: single angle Abel inversion, two angle approach, and multiple angle 2D plasma tomography. These methods were tested on symmetric and asymmetric sample functions and on experimental results from a supersonic flowing argon microwave plasma sustained in a cylindrical quartz cavity. The analysis focused on the afterglow region of the microwave flow where a plasmoid-like formation was observed. We investigated the effects of the uniform random noise added to the simulated data by applying smoothing techniques. The quality of reconstructed images was assessed by using peak signal-to-noise ratio and universal quality image measures. The results showed that the Abel inversion approach could be employed only when the system is radially symmetric, while the systems with slight asymmetry could be reconstructed with the two angle approach. In the complete absence of symmetry, full 2D tomographic reconstruction should be applied. The data analysis showed that the best results were obtained by employing either the filtered back projection or the simultaneous algebraic reconstruction technique. The total variation minimization method proved to be the best denoising technique. Each approach was used to obtain the spatial distributions of argon excited states taken at three positions along the plasmoid-like structure. The results indicated that the plasma was asymmetric with argon populating the cavity surface.

Analysis of (7)Be behaviour in the air by using a multilayer perceptron neural network.

A multilayer perceptron artificial neural network (ANN) model for the prediction of the (7)Be behaviour in the air as the function of meteorological parameters was developed. The model was optimized and tested using (7)Be activity concentrations obtained by standard gamma-ray spectrometric analysis of air samples collected in Belgrade (Serbia) during 2009-2011 and meteorological data for the same period. Good correlation (r = 0.91) between experimental values of (7)Be activity concentrations and those predicted by ANN was obtained. The good performance of the model in prediction of (7)Be activity concentrations could provide basis for construction of models which would forecast behaviour of other airborne radionuclides.