This paper using the global ionospheric map data from IGS and then using principal component analysis method and the sliding inter quartile range method to investi‐gate precursors for 4 earthquakes of Richter magnitude scale Mw ≥ 7.0 (Wenchuan earth‐quake ,Yushu earthquake ,Japan 311 earthquake ,Yutian earthquake) .Results shows that both methods can effectively detect pre‐earthquake T EC anomaly and the anomaly has stable time scale .Compared two methods ,PCA is better ,it detects the anomaly more intuitive and concise ,but the sliding inter quartile range method detects more abnormal days than PCA . However ,the results of two methods has a common trend ,and the results of two methods is consistent with the results from other scholars .In addition ,PCA is not affected by solar ge‐omagnetic activity ,this finding verify that PCA can be used in pre‐earthquake ionospheric a‐nomaly analysis .So it obvious that combine two methods effectively when analysis the pre‐earthquake ionospheric anomaly can have a better effect .%通过利用IGS提供的全球电离层格网数据对四次震级M w＞7畅0的地震（汶川大地震，玉树地震，东日本大地震，于田地震）分别利用主成分分析（PCA ）和滑动四分位法进行震前的T EC异常探测。结果显示，两种方法都能有效探测到震前T EC异常，且异常有稳定的时间尺度。两种方法相比，PC A方法更优，其探测到的异常比较直观、简明，而滑动四分位法则探测到的异常天数较多，但两者总体趋势一致，且两种方法得到的结果与其他学者分析得到的结果也一致。从探测的效果上来看PC A不易受太阳地磁活动影响，这也表明PC A的确可用于震前T EC异常分析，分析时将两种方法有效结合，相互补充，能对震前的电离层异常分析起到更好的效果。
展开▼
