It is a challenge to detect pre-seismic anomalies by using only one dataset due to the complexity of earthquakes. Therefore, it is a promising direction to use multiparameteric data. The earthquake cross partial multi-view data fusion approach (EQ-CPM) is proposed in this paper. By using this method, electromagnetic data and seismicity indicators are fused. This approach tolerates the absence of data and complements the missing part in fusion. First, the effectiveness of seismicity indicators and electromagnetic data was validated through two earthquake case studies. Then, four machine learning algorithms were applied to detect pre-seismic anomalies by using the fused data and two original datasets. The results show that the fused data provided better performance than the single-modal data. In the Matthews correlation coefficient index, the results of our method showed an 8% improvement compared with the latest study.

The sea-land telluric current vector and its continuity during two geomagnetic storms are discussed using observation data of 13 geoelectric field stations within 100 km of the coastline in China and 5 similar stations in Canada. The results show that the amplitude of the geoelectric field varies up to 300-2600 mV/km at high latitudes and in the range 100-300 mV/km at low latitudes, below 100 mV/km at the middle latitudes, when the two geoelectric storms loading. The energy spectra of geoelectric field at RES in Canada is found to be concentrated in 16-48 min, and at CHL in China concentrated in 64-128 min. The telluric current flows directly to sea from the coastal land, except the two land-type stations with particular electrical structures. The sea-land current continuity model which was set up based on geological and geophysical data to deduce vertical circulation of current channels in continental and marginal seas area, which can explain the different current orientations in different regions. Our detailed analysis show that the direction of telluric current of Island stations is related to the deposition and river erosion and controlled by ocean currents during monsoon, are also included by the model. At last, the sea-land telluric current continuity model provides well understanding for the constraints of conductivity on the sealand interface.

In order to minimize interruptions to recording, geomagnetic observatories usually use a back-up instrument operating simultaneously with the primary instrument in order to obtain comparative observations. Based on the correction parameter calculation method established in the previous work, we focused on the effects of temperature and instrument drift on the comparative geomagnetic vector observations. The linear influence of temperature on the comparative data was shown to be variable. The relative temperature coefficient changed around the temperature inflection point and showed a V-type distribution in a scatter plot. This conclusion was verified in laboratory experiments. The long-term time drift between the comparative instruments exhibits a linear pattern, and the fitness of the correction model can be evaluated by the degree to which the residual distribution of the fitted straight line conforms to the normal distribution. However, the absolute value of the long-term time drift between variometers with the same type of probe is very small. Therefore, long-term time drift correction should be carried out with care. The associated analysis and conclusions have the potential to benefit data agreement correction of long-term comparative geomagnetic vector observations and comparative testing of the performance of vector instruments.

Aiming at the problems of huge storage space, low exchange speed and low read-write speed of the current specific oracle database, the read-write speed and exchange speed tests are performed on the compressed and uncompressed Clob and Blob data by three compression algorithms, including Bzip2, Gzip and GzipIO respectively. The read speed test is performed by the direct read, substr read, and substr+threadPool read techniques. The results show that: (1) Blob is superior to Clob in terms of storage, exchange, or read-write speed; (2) For the specific database, Blob+Gzip is the optimal storage structure of the minute and second data. The read-write speed is greatly improved, and the overall capacity of the database is reduced to 7% (or less). The exchange rate of the second data is at least 7.89 times of the present rate, and the station data can be exchanged to the disciplinary center within 2-3 hours (currently 1.5 days); (3) The simplest and most widely used direct read method by software developers has poor database read efficiency, while the substr+threadPool technique shows higher database read efficiency no matter for Clob or Blob, for compressed or uncompressed, which brings a leap-forward improvement in the read speed of LOB data. The results of this paper are of high reference significance to the LOB data storage design and software development.

The singular spectrum analysis (SSA) method has the characteristics that it is not affected by the noise spectrum distribution, and is superior to the traditional denoising method. Moreover, it has been commonly used in signal analysis in the fields of oceanography, machinery, and electronic technology. This paper systematically introduces the basic principle and implementation process of singular spectrum analysis method. Then the singular spectrum analysis method is applied to the interference identification and removal of seismic precursor observation data. Taking deformation data and geoelectric field observation data as an example, the results show that the singular spectrum analysis method can effectively separate the rainfall interference in deformation observation data, and has important guiding significance for the subsequent removal of rainfall interference. In addition, the singular spectrum analysis method can be used to extract and remove the interference of HVDC and subway from the geoelectric field observation data, and the effect is obvious, which can effectively improve the data quality and better serve the earthquake analysis and prediction.