An algorithm to remove fringe jumps and its application to microwave reflectometry

In some plasma discharges, the phase measured by microwave reflectometry has many fringe (2 pi radians) jumps. A new algorithm to detect and remove fringe jumps has been developed, and applied to the data in the JIPP TII-U tokamak. Using this algorithm, quantitative properties of fringe jumps, and their effects on the analysis of phase fluctuations are investigated. It was found that the occurrence of fringe jumps obeys a Poisson process, and the time scale of jumps is distributed over a wide range. Fringe jumps affect mainly the low-frequency components of phase fluctuations. Comparison of the phase corrected by the algorithm and the phase calculated from the time smoothed signals indicates that time smoothing (or frequency filtering) is an effective way to obtain information concerning the macroscopic density profile. Fringe jump and phase runaway can be phenomenologically explained by the distribution of the complex amplitude of the reflected wave.