Soft Computing Technique for Recognition of Earthquake Precursor from Low Latitude Total Electron Content (TEC) Profiles

Abstract

Total Electron Content (TEC) data from GPS are now used as tools for identifying an impending earthquake. The extraction of earthquake induced features from this parameter needs elaborate processing because it involves filtration of data with respect to disturbed day variations, contribution from multi path effects and also normal day-to-day fluctuations even during quiet days. In this paper we introduce, a peak detection algorithm, an automatic pattern matching approach and artificial neural network for processing GPS generated TEC record in where a template is framed from the time series of quiet day data for extraction of pre -seismic parameter. This matching algorithm is adopted to find deviation in TEC time series prior to an impending earthquake from the template and would be used as an index of earthquake induced signature. A few case studies using these techniques are presented in this paper. Work is based on TEC data collected from GPS at Gauhati (26º 10' N, 91º 45' E).

Detecting All Possible Ionospheric Precursors by Kernel-Based Two-Dimensional Principal Component Analysis

“…TEC above the forthcoming epicenter was found to anomalously decrease in the afternoon period in the fourth to sixth days and in late-evening period on the third day before the earthquake but was enhanced in the afternoon in the third day before the earthquake. Reference [19] adopted a soft computing technique to identify TEC anomalies related to earthquakes with low latitude TEC data using the peak-detection technique and dynamic time-warping algorithm for pattern reorganization. This is the first step in utilizing this approach for identification of an impending earthquake.…”

“…Moreover, a lot of researchers need to process TEC data. (4) The mathematical algorithms, e.g., dynamic time-warping algorithm [19], were only used to distinguish earthquake-related TEC anomalies from other unknown TEC anomalies. (5)The quality of the analysis, researching observing TEC variants as TEC precursors, was always affected by space weather conditions.…”

“…A TEC unit is TECu (10 16 electrons/m 2 ) [16]. Ionospheric TEC precursors of large earthquakes have been widely researched both as precursors and after effects ( [43], [45], [62], [31], [19], [12], [20], [46], [38], [49], [52], [1], [24], [40]). Ionospheric TEC precursors of earthquakes are generally of two types: physical and chemical precursors.…”

Predicting Ionospheric Precursors Before Large Earthquakes Using Neural Network Computing and the Potential Development of an Earthquake Early Warning System

“…Several researchers have already applied the concept of AI to detect TEC anomalies associated with earthquakes. Kalita et al (2012) used a peak detection algorithm, an automatic pattern matching approach, and an artificial neural network (ANN) to process GPS data, generate the TEC record, and detect the TEC precursor of an impending earthquake in Guwahati, India, with an epicenter at 26.10° N and 91.45° E. Homam (2016) presented TEC anomalies related to earthquakes using an ANN, GPS ionospheric scintillation, and a TEC monitor (GISTM) receiver from 2005 to 2010 during low-to medium solar activity, with sunspot numbers (SSNs) of 0.0-42.6. To reduce the predicted errors, Okoh et al (2016) used a type of ANN model for the Global Navigation Satellite System (GNSS) and the vertical TEC (GNSS-VTEC) in Nigeria.…”