C_（n）及其在异常震群判别研究中应用的初步尝试

C(N) AND ITS PRIMARY APPLICATION IN STUDIES OF DISTINGUISHING ANOMALOUS SWARMS

本文引入算法复杂性度量Ｃ（ｎ）判别震群活动是否具有前兆意义。选取１９７１－１９９３年间３．２≤Ｍｍａ≤ｘ５．０，Ｎ≥１２的震群序列，用１０小时段，五种震级状态计算Ｃ（ｎ），发现当Ｃ（ｎ）≥８时，与其后１６４天内四川及西南部周边附近５级以上强震相关。Ｒ评分具有９７．５％置信水平。同时，分析、比较了Ｃ（ｎ）与参数ｂ、Ｋ、Ｕ、Ｈ的预报能力，认为Ｃ（ｎ）有一定的预报效能且高于Ｈ、Ｕ和Ｋ。

Algorithm complexity C(n) was introduced in this paper to identifywhether the swarm activity makes precursory sense. Swarm sequences withN≥12 and 3.2≤Mmax≤5.0 from 1971 to 1993 are used, C(n) was calculatedby 5 magnitude states With a step of 10 hours. It is found that C(n)≥6is correlated with strong earthquakes of M≥5 in 164 days occurred inSichuang and neighboring areas. Reliability reaches as many as 97.5%.Meanwhile, the prediction abilities of C(n) and parameters b, K, U, Hhave been analysed and compared that C(n) has a certain prediction efficacyhigher than H, U and K.