On the Correction of Spatial and Statistical Uncertainties in Systematic Measurements of 222Rn for Earthquake Prediction

Abstract

In earthquake prediction studies, the regional behaviour of accurate 222Rn measurements at a set of sites plays a significant role. Here, measurements are obtained using active and passive radon detector systems in an earthquake-active region of Turkey. Two new methods are proposed to explain the spatial behaviours and the statistical uncertainties in the 222Rn emission measurements along fault lines in relation to earthquake occurrence. The absolute point cumulative semivariogram (APCSV) and perturbation method (PM) help to depict the spatial distribution patterns of 222Rn in addition to the joint effects of the K dr, the radon distribution coefficient, and the perturbation radon distribution coefficient (PRDC). The K dr coefficient assists in identifying the spatial distributional behaviour in 222Rn concentrations and their migration along the Earth’s surface layers. The PRDC considers not only the arithmetic averages but also the variances (or standard deviations) and the correlation coefficients, in addition to the size of the error among the 222Rn measurements. The applications of these methodologies are performed for 13,000 222Rn measurements that are deemed to be sufficient for the characterization of tectonics in the Keban Reservoir along the East Anatolian Fault System (EAFS) in Turkey. The results are evaluated for the Icme earthquake (M L 5.4, 5.7 km, 23 June 2011), which occurred in the vicinity of the EAFS.