PRELIMINARY STUDY ON MULTI-LEVEL EARTHQUAKE PREDICTION BASED ON GROUNDWATER DATA

On basis of dynamic groundwater data in Northern China during the past 40years,repeatability abnormal information represented characteristic of different time scales are extracted from data of monthly mean,daily mean and minutely mean respectively for the studying of multi-level earthquake prediction. The results show that before strong earthquakes,the groundwater in Northern China has a 10-year quasi period;the raising and high value segment of the quasi period may be a long trend abnormal change,in the raising and high value segment there are medium-short 3-year declining abnormal changes,short and impending complex abnormality in different forms appears form 1day to several months.These obvious repeatable abnormalities in different time scales can be used for multi-level earthquake prediction to strong earthquakes.