Detection of atmospheric radon concentration anomalies and their potential for earthquake prediction using Random Forest analysis

Mayu Tsuchiya,Hiroyuki Nagahama,Jun Muto,Mitsuhiro Hirano,Yumi Yasuoka

DOI: https://doi.org/10.1038/s41598-024-61887-6

IF: 4.6

2024-06-01

Scientific Reports

Abstract:Various anomalies occurring before earthquakes are currently being studied to predict seismic events, with one of them being the radioactive element radon ( 222 Rn). Radon concentrations in the soil, water, and atmosphere fluctuate in response to crustal movement. Recent research has statistically detected anomalies by analyzing the fluctuations in radon concentrations before earthquakes and conducting quantitative evaluations of radon. However, the method used to determine the parameters in the analysis was problematic. Therefore, in this study, we compared observed atmospheric radon concentration data with predicted values based on typical annual patterns using Random Forest analysis. We conducted a more objective analysis by employing this method and statistically determining anomalies using thresholds. This analysis was conducted using atmospheric radon concentration observation data obtained at Kobe Pharmaceutical University (KPU) before the 1995 Kobe Earthquake, and ionization currents emitted when radon decays were obtained at Fukushima Medical University (FMU) before the 2011 Tohoku-oki Earthquake. Consequently, before the major earthquakes occurred at both locations, the difference between the predicted and observed values exceeded the standard deviation by a factor of three. These results indicate the potential of Random Forest analysis to identify anomalies in atmospheric radon concentrations before earthquakes occur.

multidisciplinary sciences

What problem does this paper attempt to address?

The main objective of this paper is to explore the possibility of using the Random Forest analysis method to detect abnormal changes in atmospheric radon concentration before earthquakes and its potential in earthquake prediction.

The paper first introduces radon (222Rn) as a radioactive element and discusses the changes in its concentration in soil, water, and the atmosphere before earthquakes, as well as the correlation between radon concentration and crustal movements. It then points out the issues in previous studies, specifically the problems with the methods used to analyze radon concentration anomalies, particularly the uncertainty in parameter determination, which leads to ambiguity in anomaly detection.

To address these issues, the authors employed the Random Forest analysis method to compare the observed atmospheric radon concentration data before earthquakes with data predicted based on typical year patterns. This method is more objective and determines abnormal changes through statistical thresholds. Specifically, they used atmospheric radon concentration data collected at Kobe Pharmaceutical University (KPU) before the 1995 Hanshin earthquake and ion current data collected at Fukushima Medical University (FMU) before the 2011 East Japan earthquake.

The research results show that the differences between observed values and predicted values exceeded three standard deviations before the major earthquakes at both locations, indicating that the Random Forest analysis can effectively identify abnormal changes in atmospheric radon concentration before earthquakes. Additionally, the paper discusses the challenges faced by actual monitoring networks, including the need to verify the method's effectiveness in different tectonic environments and consider the impact of solar activity.

In summary, the issues this paper attempts to address are:
- How to more objectively detect abnormal changes in atmospheric radon concentration before earthquakes.
- Exploring the potential application of the Random Forest analysis method in earthquake prediction.