A persistent radon anomaly signal preceding the destructive 7.7 Mw earthquake in Myanmar on March 28, 2025

Abstract

On March 28, 2025, a devastating earthquake doublet of moment magnitude 7.7 and 6.7 struck the Sagaing Region of Myanmar, causing extensive damage and significant casualties across Southeast Asia. In the months preceding this seismic event, a continuous radon monitoring system-BhaROSA-installed in Imphal, India, as a part of the Indian Network for Detecting Radon Anomaly signal (INDRA), recorded a pronounced and statistically significant novel radon anomaly signal. The anomaly signal commenced on December 5, 2024, with a gradual buildup, followed by a sharp rise on February 28, 2025. Multiple peak alerts were observed prior to the mainshock (7.7 M_w) on March 28, 2025, after which the signal declined rapidly, returning to baseline levels-suggesting a potential correlation with pre-seismic crustal stress accumulation and release. The anomaly signal exhibited a normalized squared deviation of 35.14 from the baseline value, far exceeding natural variability at the time of main shock. The radon signal, with a build-up period of approximately 109 days and a decay of ~ 96 h, closely matched the spatial and temporal characteristics of the dilatancy-diffusion model. A pooled analysis of radon anomalies from ten earthquake events, including this major event, across multiple Indian observatories of INDRA reveals a robust positive correlation (r = 0.96, R² = 0.93) between radon buildup duration and earthquake magnitude. These findings strongly suggest that radon emissions are sensitive indicators of impending seismic activity and radon build up period can be a good indicator of magnitude of earthquake. The study highlights the potential of continuous radon monitoring in tectonically active regions like Northeast India and Myanmar as a viable component of earthquake precursor research and early warning systems.

Keywords: BhaROSA; Earthquake precursor; Myanmar earthquake 2025; Radon anomaly; Sagaing fault.

Fig. 1

Macro-seismic Intensity Map of Myanmar earthquake 7.7 M_w that occurred on March 28, 2025, at 06:20:52 UTC (Big Data Viz (Ver 0.1) Interface, Ellyptech, India, URL: www.ellyptech.com/init/products/bigdataviz).

Fig. 2

Schematic of Measurement set up of Bhabha Radon Observatory for Seismic Application (BhaROSA) deployed at Imphal, Manipur, India.

Fig. 3

Location of Radon Observatory and 3 earthquake events that occurred in the vicinity of the observatory during radon anomaly period (Big Data Viz (Ver 0.1) Interface, Ellyptech, India, URL: www.ellyptech.com/init/products/bigdataviz).

Fig. 4

Time Series Variation of Radon Counts with Ambient Temperature, Relative Humidity, and Atmospheric Pressure at Imphal Radon Observatory.

Fig. 5

Violin box plot of Radon, temperature, Rel. Humidity and Pressure measured at Imphal during June 07, 2024 to April 11, 2025.

Fig. 6

Time series plot of radon data at Imphal Observatory showing (from bottom to top): raw radon counts, the extracted diurnal component, and the residual radon signal after removal of diurnal and high-frequency noise. This decomposition highlights the underlying long-term anomaly trend, distinct from natural diurnal variations, and facilitates clearer identification of geophysically significant deviations.

Fig. 7

Time series plot showing the relative changes in radon, temperature, humidity, and atmospheric pressure during the radon anomaly progression at Imphal Observatory. While radon exhibits a significant and sustained deviation, no corresponding anomalies are observed in the meteorological parameters. The stability of temperature, humidity, and pressure throughout the period effectively rules out meteorological influences as the cause of the observed radon anomaly.

Fig. 8

Time series of radon counts (Daily MovAvg, Monthly MovAvg) at Imphal observatory showing precursor anomaly prior to the main event , 7.7 M_w Myanmar Earthquake.

Fig. 9

Dilatancy and micro-cracking model to explain radon anomaly signal prior to seismic event.

Fig. 10

Correlation between earthquake magnitude and radon signal buildup period with epicentral distance scaling.

Fig. 11

The metric of true and false positive along with detected and missed seismic events. An anomaly is classified as a True Positive if an earthquake occurs within 100 h and has an epicentral distance less than the strain radius of the earthquake; otherwise, it is classified as a False Positive. A detected event is an earthquake that occurs with at least one prior radon anomaly within 100 h from an observatory located within the strain radius, whereas a missed event is an earthquake that occurs without any prior radon anomalies.