Mulatijiang Abulaiti, JIN Hua, TENG Hai-tao, MAO Yu-jian
Application of wavelet analysis and correlation analysis method, the quantitative study of meteorological elements (temperature, air temperature, air pressure), water level and the elm ditch 4 set of observations, the relationship between the oblique observation instrument and regression equations of each component were obtained by regression analysis, through the original curve and regression residual sequence contrast, summarizes the anomaly characteristics and earthquake cases, The R value scoring method is used to test the seismic reflection efficiency. The results show that: (1) The dominant period of each component of the barometric pressure on the Yushugou tilt observation is 4 069~8 192 min, which is mainly linear in this frequency band; (2) The ground temperature is the main influencing factor of the annual variation information of Yushugou tilt observation (except for the EW direction of horizontal pendulum and the NS direction of borehole tilt), and has a linear relationship with the observation value; (3) Water level is the main influencing factor of the annual variation information of the horizontal pendulum in NS direction and the water pipe gauge in EW direction, and has a linear relationship with the observed value; (4) Regression analysis can effectively eliminate the influence of meteorological factors and water level to a certain extent, and has amplification effect on the identification of seismic anomalies; (5) The result of R-value test shows that the best prediction window of EW component of horizontal pendulum is 270 days, the best R=0.54, R0=0.41, and the best prediction period of NS component of water meter is 340 days, the best R=0.44, R0=0.40, all of which have passed the prediction efficiency test, indicating that they have high abnormal reliability.