ARIMA model simulation for total electron content, earthquake and radon relationship identification

Abstract

Earthquakes cause many losses of life and property with their devastating effects. Scientists conduct studies to predict the hazards by examining the anomalies that occur before the earthquake. In this study, mathematical and statistical relationships are examined between soil radon (Rn-222) gas and earthquake and atmospheric total electron content (TEC). Furthermore, an Autoregressive Integrated Moving Average (ARIMA) simulation model is proposed to predict Rn concentrations. The model is evaluated for the M 4.2 Sivas, Susehri earthquake in Turkiye that took place on the North Anatolian Fault Zone in 2007 and a relationship is determined between soil Rn gas and micro-seismic activity. In parallel with the earthquake-radon relationship, some meteorological variables [5, 10, 20, 50 cm soil temperature (degrees C), vapour pressure (hPa), wet bulb temperature, dry bulb temperature] are identified as associated with the earthquake. It is also observed that the TEC increases with the relative Rn gas concentration as the time of the main shock is approached. This provides meaningful results for further seismo-ionospheric change interpretations. In addition, the ARIMA model detects possible future Rn gas concentration values. [GRAPHICS] .