Drastic changes before the 2011 Tohoku earthquake, revealed by exploratory data analysis

Predicting earthquakes is critical, particularly in high-risk nations, and despite significant efforts, it has yet to be realized. Nonetheless, there is a scarcity of statistical methodologies in seismic investigations, to the point that an old idea is accepted without testing. Exploratory data analysis (EDA) was used to examine seismic recordings in Japan in terms of magnitude and timing. EDA is a parametric statistical method that uses data features to conduct data-driven investigations. The distribution style of each dataset was established, and the key factors were identified. This allowed us to detect and assess irregularities in the data. Swarm earthquakes occurred at very high frequencies before to the massive 2011 Tohoku earthquake.

Download (12.18 kB)

Seismic Pattern Changes Before the 2011 Tohoku Earthquake Revealed by exploratory data analysis

The capacity to forecast earthquakes is extremely beneficial, especially in high-risk seismic zones, yet exact predictions remain difficult. One possible explanation is the insufficient incorporation of statistical techniques into earthquake research. In this work, I used exploratory data analysis (EDA), a data-driven parametric statistical approach, to look at earthquake records from Japan, utilizing data given by the Japan Meteorological Agency. The intervals between earthquakes closely matched an exponential distribution, given by a single parameter, λ, indicating occurrence frequency. Unlike the standard Gutenberg-Richter law, earthquake magnitudes follow a normal distribution with two parameters: µ (mean) and σ (scale). After creating these distributions and their parameters, considerable changes became apparent.

Download all (1.01 MB)