【摘要】 对华北地区１７ 个台站２５ 套倾斜仪１０ 年以上的资料，统一用调和分析、卡尔曼滤波和周期滤波方法进行全面系统处理，提取异常信息，统计分析与该地区ＭＳ５ 级以上地震的关系，并对这几种方法进行评价。结果是：地倾斜固体潮Ｍ２ 波γ因子异常与地震的对应率为５７％ ，Ｒ＝０．４０；调和分析残差异常对应率为４１％ ，Ｒ＝ ０．２５；由卡尔曼滤波方法识别的异常与地震的对应率为１５％ ，Ｒ＝ ０．１１；周期滤波方法识别的异常与地震对应率为５９％ ，Ｒ＝ ０．５１。建立了地形变预报地震的综合模型，异常信度在空间上的分布呈同心圆状，地震通常发生在信度高的地区。更多还原

【Abstract】 The data of 25 sets of tiltmeter in 17 stations over 10 years in north China are processed comprehensively and systematically by the methods of harmonic analysis, Kalman filtering and periodic filtering. The anomaly information of these data and the relation between anomaly and earthquakes ( M S>5) in the region is analyzed statistically. In this way, all methods are evaluated. The results show that 57% anomalies ( R =0.40) identified from γ factor of M 2 waves of ground tilt solid tides, 41% anomalies ( R=0.25 ) picked up from harmonic analysis and 15% anomalies ( R =0.51) recognized from periodic filtering are corresponded to earthquakes. A synthetic prediction model with crustal deformation approach is established. The distribution of anomaly confidence shows elliptic shape, and earthquake usually occurs in high confidence value region.更多还原