The data of the article "Pulsar Time and Atomic Frequency standards Combined Time-scale"

This dataset mainly includes clock bias data of integrated pulsar time EPT62 and international atomic time TAI (EPT62-TAI), EPT62-TAI smoothed with a combined smoothing filter, and atomic frequency reference and TAI frequency difference data smoothed with a smoothing filter. In the study, the EPT62-TAI clock bias time series was calculated using the timing observation data of 62 millisecond pulsars from Version A, which was publicly released for the second time by the International Pulsar Timing Array (IPTA). The total time span of EPT62-TAI data is 29 years. Using a combined smoothing filter, the EPT62-TAI clock difference time series, atomic frequency reference, and TAI frequency difference time series are smoothed to obtain the smoothed EPT62-TAI, atomic frequency reference, and TAI frequency difference. The smoothed EPT62-TAI is also known as CPA-TAI. According to the principle of combined smoothing filter, the frequency difference of CPA-TAI is equal to the frequency difference between the atomic frequency reference smoothed by the combined smoothing filter and TAI. The frequency stability and accuracy of CPA are very close to the widely used Earth time, therefore, CPA can also be used as Earth time. This dataset mainly consists of two files. File 1 is the comprehensive pulsar time and TAI clock bias (EPT62-TAI) data ("clock62"), as shown by the blue line in Figure 1 of this article (60 rows and 3 columns): the first column is the reduced Julian day (MJD), the second column is the clock bias (unit: μs), and the third column is the uncertainty of the clock bias (unit: μs). File 2 is the data smoothed with a smoothing filter ("clock62cp") (382 lines and 3 columns): the first column is the reduced Julian day (MJD), the second column is the CAP-TAI clock bias (unit: μs), as shown by the red line in Figure 6 of this article, and the third column is the frequency difference of CPA-TAI (unit: μ s/d), as shown by the red line in Figure 5 of this article. Figures 4 and 7 in the article are calculated using data from files 1 and 2, with the aim of evaluating the performance of these two subsets of data. Therefore, relevant data will not be provided.