Abstract: This paper describes the basic statistical characteristics of non-stationary Poisson process of the events and differences between non-stationary and stationary Poisson processes, and applies the non-stationary Poisson process to study the long-and medium-term probabilistic earthquake prediction in North China seismic area. Suppo sing that the non-stationary Poisson time process of strong earthquakes ( M_S≥7.0) , in the last seismic cycle for the Fenwei and North China Plain seismic zones in North China seismic area would recur in the coming several hundred years, the cumulative probability for occurrence of a strong earthquake (M_S≥7.0) in the two zones should respectively be 0.26( uncertainty range: 0.06~0.50) and 0.04( uncertainty range: 0.00~0.13) before 2010, which are obviously less than that calculated based on the stationary Poisson process.

Key words: Non-stationary process, Long-and Medium-term probabilistic earthquake prediction, North China seismic area