由中国古代月掩犯记录得到的地球自转长期变化

SECULAR VARIATION OF EARTH'S ROTATION FROM ANCIENT CHINESE RECORDS OF LUNAR OCCULTATIONS AND APPROACHES

本文作者发展了一种"时间窗"(实际可见时间段)方法,以利用中国古代无计时的月掩星记录研究地球自转长期变化,并对中国古代月掩星和月犯星资料进行了全面统计分析.在此基础上,本文由五代以前的732条月掩犯记录进行了地球自转分析,这一工作填补了公元三至七世纪的资料弱环,得到的地球自转加速度平均值与前人的结果相当一致。按时间分段得到的结果分别与前后衔接并显示出明显的变化发生于公元五世纪.利用元代240条月掩犯记录的分析证实,"时间窗"方法没有明显的系统误差。

In recent years the study of secular variation of earth's rotation has got great achievement.Ancient records not only proved the hypothesis of tidal dissipation, but also indicated another nontidal term which is accelerating the earth's rotation. The best materials gather in two periods: 7th century BC to 2nd century AD and 9th to 12th century AD. Since results from the two parts are apparently. different, the weak link of ancient records from 3rd to 8th century is very noticeable [12]. Just in those periods, there were very many records of occultations and approaches by the moon of both planets and stars contained in chinese texts which may remedy the weak link although these untimed records are quite rough. The author developed a new method called 'Time Window (Really Visible Time)' in order to improve these untimed ancient observations. Primary experiments were successful +I1 2] . In another work [3] , all of these records are fully collected, examined and analyzed. In this paper, 732 records of lunar occultations and approaches before 960 AD (mainly from third to seventh century) are chosen to study the earth's rotation after a detailed discussion to accuracy and other characteristics of the records. The obtained mean acceleration ω is quite consistent with other sources. The records are divided into several groups according to time order and the results are computed for each group. It is interesting that former groups agree with the first part mentioned above, while later groups coincide with the other part. An apparent change took place in fifth century. The fascinating disoovery needs further proof and investigation. Two hundred and forty similar records in 12th century are also analyzed. The result shows that the method 'Time Window' has no any significant systematic error. This is an important conclusion.