Abstract: 12 digital observation stations of water level have been set up in Gansu and its adjacent areas. The Pingliang C11 well is uniquely observed with digital data and simulation at the same time. We analyze dynamic characteristics of year, month and day digital water level and application of digital data. The interfere factors that affect digital data of water level are also analyzed. The results show that prominent problem is sudden change and pulse in digital data of water level. According to summation of earthquake cases, we can identify the intermediate-term precursory anomaly in breaking the annual variation. It is difficult to identify short time precursory anomaly in sudden change and pulse. We processed the simulation water level data by using the tide theory, and indicating that this theory can be applied to process the digital observation data.

Key words: Digital observation data, Underground water level, Dynamic characteristics, Gansu and its adjacent areas, Earthquake prediction efficacy