Study on extracting medium－term precursory anomaliesfrom water radon and water level observationwith wavelet analysis

Noticing that the wavelet analysis possesses a characteristic of high resolution in time and frequency domain, this paper develops a method of extracting wavelet medium-term anomaly from water radon and water level observation.The extraction process includes,through research, deciding the objective earthquake expected to be predicted, selecting testing items and observation wells, extracting wavelet anomaly, identifying the property of the precursor anomaly, etc.(seven steps). With the methodology of extraction of wavelet anomaly, we have studied large amount of earthquake materials in North China, Sichuan and its vicinity,and Hainan. After analyzing the relation between the earthquakes and the anomaly of water level and radon, we provided the characteristic time distribution of wavelet anomalies in these three regions. Moreover, the advanced time of the wavelet anomaly can be regarded as an indicator for forecasting an earthquake.This indicator would be helpful in analyzing earthquake situation in practice.