Study On Short-term Forecasting Of The Ionospheric Precursors Of Earthquake

Earthquakes (EQ) are the most hazardous natural disasters in human history, and how to predict EQ is a worldwide problem. Because of EQ is a physical process of accumulation and release of enormous energy in the crustal movement changes, accompanied the development of the EQ, a series of geophysical effects may occur around the seismic source spatial area, which resulted some anomalies in many aspects, such as stress-deformation, hydro-meteorological, geological chemical and electromagnetic disturbance, we called earthquake precursors. Ionospheric precursors of EQ, a new short-term EQ precursor by a growing concern, which is explored in the recent years, developed a new way to predict EQ. In this paper, we use ionospheric data from oblique and vertical ionosonde networks over China, developed a new method for predict ionospheric precursors of EQ. The main research results are presented as follows:First, analysis of foF2changes in China region. We use the historical observed data in China region to analyze ionosphere in quiet time and in ionospheric storm during2000in China region; and the correlation between ionosphere and solar in China region, and the seasonal response of the ionosphere in the magnetic storms in2000; then summarized the characteristics of the ionospheric storm. On this basis, we introduce the ionospheric precursors of EQ and analyze the similarities and differences with the ionospheric storm.Second, we build a one to three days ahead of forecast ionospheric disturbances model in China region. Based on the analysis of the correlation between ionospheric disturbances and solar activity, we use neural network (NN) to build a model for1to3days ahead of forecast ionospheric disturbances in China region. The results of comparing with actual data indicate that this model is very promising.Third, we developed a regional reference ionosphere of China region. A regional reference ionosphere model of China region is developed by neural networks (NNs) trained by genetic algorithm (GA). In order to avoid the’local minimum’phenomena in most NN applications, GA is utilized here to optimize the initial weights of NNs. The input parameters used in this GA-NN based foF2prediction model consist of BJT (GMT+8), day number (day of the year), seasonal information, solar cycle information, magnetic activity, magnetic declination, magnetic dip angle, angle of meridian relative to sub-solar point, solar zenith angle and geographic coordinates. Prediction results of GA-NN model, unimproved NN model and International Reference Ionosphere2007model (IRI2007) are compared with the observation data of1996and2000respectively. The results indicate that the predicting accuracy of GA-NN model is improved by about10%than the IRI2007.Fourth,1to24hours ahead predicting of ionospheric storm for single station in China region. An1to24hours ahead predicting of ionospheric storm model for single station in China region is developed by neural networks (NNs) trained by genetic algorithm (GA). The results show that the model is very promising.Fifth, we developed a TEC regional extrapolation model in East Asian. Because of the current common reconstructions of ionospheric regions are only interpolations in the region, we developed a TEC regional extrapolation model in East Asian sector using the data from IGS to be a simple attempt. By comparison with the measured data to prove that the method is reliable and effective, and shows a good prospect.Sixth, a real-time mapping model offoF2in north China has been established by neural networks (NNs) improved by GA based on the oblique networks data. The model can real-time map in north China, and shows a good prospect of application in engineering applications.At last, we developed a preliminary model for short-term forecastion and analysis of ionospheric precursors of EQ based onfoF2real-time mapping model and regional reference ionosphere of China region. This model is able to determine whether the ionosphere is anomaly by comparing the real-time map and regional reference ionosphere. If the ionosphere is anomaly, the model can be a detailed analysis of the ionospheric anomalies then determine whether it is ionospheric precursors of EQ. At last we develop software by the model.