LMD and Lomb–Scargle power spectrum variability analysis of quasars

The observation of celestial light variation data is discontinuous and uneven in the time domain due to weather, equipment, and other factors. And the discontinuous observation data introduce a significant error in the study of the varying light periods. This paper combines the newly developed local mean decomposition(LMD) method and the Lomb–Scargle (LS) algorithm, the most widely used period estimation algorithm for non-uniform time series; a method for analizing the optical variable period of quasars based on the LMD and the LS algorithm is proposed. The principle and the application steps of the method were discussed. The observation data for OJ 287 and 3C 345 at 4.8, 8.0, and 14.5 GHz (radio bands) were collected and analyzed from the University of Michigan UMRAO database. For OJ 287, periods of 12 ± 0.3, 5 ± 0.65, and 2.4 ± 0.01 yr are found in all three bands. There also exist periods of 11 ± 0.9, 5 ± 0.31, and 3 ± 0.28 yr in the three bands for 3C 345. These periods agree with the results of the methods commonly used in some literature. At the same time, some other possible periods are also obtained that can provide references for further research. Results show that estimating the power spectrum is feasible based on LMD+Lomb–Scargle, which provides a new idea for the photoperiod analysis of quasars.