Modeling of the Coseismic Electromagnetic Field Observed during the 28 September 2004, M 6.0 Parkfield Earthquake

Tuesday, 16 December 2014
Yongxin Gao1,2, Jerry M Harris1, Jian Wen2, Xiaofei Chen2 and Hengshan Hu3, (1)Stanford University, Stanford, CA, United States, (2)University of Science and Technology of China, Hefei, China, (3)Harbin Institute of Technology, Harbin, China
On 28 September 2004, the M6.0 Parkfield earthquake took place on the San Andreas fault, California. A seismic station which is named PKD and located near the epicenter recorded both of the seismic and electromagnetic (EM) signals during this earthquake. This station is operated by Berkeley Seismological Laboratory and installed with broadband seismometer and EM sensors which are close to each other. Significant seismic signals as well as clear coseismic EM signals were recorded during this earthquake, providing a good opportunity to study the coseismic EM phenomenon.

We modeled the coseismic EM signals from the viewpoint of the electrokinetic effect on the basis of Pride‚Äôs equations. The earthquake source is taken as a finite fault with length of 40 km along the strike direction and width of 15 km along the dip direction. The source parameters that we use for calculation were inverted by Liu et al. [2006, BSSA] by utilizing the seismic data. While in their inversion the earth crust are treated as 7 horizontally-layered elastic solids, in our calculation these solid layers are regarded as porous media. Each porous layer has the same P-velocity, S-velocity and density to its counterpart solid layer. The salinity is set to be 0.1 mol/L for all the layers so that conductivity is uniformly distributed with the value of 0.036 S/m. To evaluate the electric and magnetic responses during the rupturing of the earthquake, we use the algorithm developed by Hu and Gao [2011, JGR] which calculates both the seismic and EM wavefields simultaneously. Since the inversion of source parameters was operated in the frequency band 0.16 Hz-1 Hz, we filter both of the synthetic seismoelectric wavefields and the real data before making comparison between them. Our preliminary result shows that in this frequency range, the amplitude of the simulated coseismic electric field is of the order of 1µV/m, which is the same to the real electric data. This supports the electrokinetic effect to be an effective and reasonable mechanism for the generation of the coseismic electric field.