Critical analysis of the electrostatic turbulence enhancements observed by DEMETER over the Sichuan region during the earthquake preparation

Abstract. In this paper, we report initial results from a detailed statistical study of the plasma waves observed by the DEMETER satellite over the Sichuan region during a period of 20 days encompassing the large earthquake of magnitude M =7.9 that occurred on 12 May 2008. The main objective of this paper is to present a statistical method to process and analyze plasma wave data and assist in detecting possible earthquake precursors among larger irregular disturbances arising from the natural variability of the ionized environment of the Earth. This method, presently used for dayside observations, involves two stages. First, VLF wave spectra are processed to recognize the various types of plasma waves usually observed at mid and low latitudes and derive a reduced number of parameters that fully characterize these emissions and may be conveniently used for a detailed statistical study. In a second stage, we perform a statistical analysis of the results by taking into account two "reference zones" displaced respectively 30^° eastward and westward from the "epicentre zone". Plasma and wave disturbances possibly induced by earthquakes in preparation are likely to maximize close to the "epicentre zone", while natural disturbances associated, in particular, with the varying magnetic activity are rather uniform over a wider longitude sector, thus enabling the use of observations over the reference zones as a base line. The initial results of this study show a deviation of the power spectrum of electrostatic turbulence in the epicentre zone about 6 days prior to the earthquake but no significant anomalous variations can be observed on other characteristics of plasma waves. From the analysis of the data over the two reference~zones and using recently produced sector magnetic activity indices, we conclude that the enhancement of electrostatic turbulence is associated with magnetospheric processes rather than with pre-seismic activity.