Standardised Radon Index (SRI): a normalisation of radon data-sets in terms of standard normal variables

R G M Crockett,C P Holt

doi:10.5194/nhess-11-1839-2011

Abstract

Abstract. During the second half of 2002, from late June to mid December, the University of Northampton Radon Research Group operated two continuous hourly-sampling radon detectors 2.25 km apart in the English East Midlands. This period included the Dudley earthquake (ML = 5, 22 September 2002) and also a smaller earthquake in the English Channel (ML = 3, 26 August 2002). Rolling/sliding windowed cross-correlation of the paired radon time-series revealed periods of simultaneous similar radon anomalies which occurred at the time of these earthquakes but at no other times during the overall radon monitoring period. Standardising the radon data in terms of probability of magnitude, analogous to the Standardised Precipitation Indices (SPIs) used in drought modelling, which effectively equalises different non-linear responses, reveals that the dissimilar relative magnitudes of the anomalies are in fact closely equiprobabilistic. Such methods could help in identifying anomalous signals in radon – and other – time-series and in evaluating their statistical significance in terms of earthquake precursory behaviour.

Highlights

There is an increasing body of evidence which indicates that radon emissions from rocks, soils, and groundwater can provide a diagnostic tool for some geophysical processes, e.g. tidal deformations and earthquakes
Standardising the radon data in terms of probability of magnitude, analogous to the Standardised Precipitation Indices (SPIs) used in drought modelling, which effectively equalises different non-linear responses, reveals that the dissimilar relative magnitudes of the anomalies are closely equiprobabilistic
We have applied moving averages to reduce the effects of drift/mismatch in sampling times, and techniques such as Empirical Mode Decomposition (EMD) to de-noise time-series and reduce the effects of different detector responses

Summary

Introduction

There is an increasing body of evidence which indicates that radon emissions from rocks, soils, and groundwater can provide a diagnostic tool for some geophysical processes, e.g. tidal deformations and earthquakes In this context, it is often informative to compare two radon data-sets, e.g. variations or anomalies in radon concentrations at different locations which might be responses to common stimuli. As a means of investigating similarity of response to stimulus, a Standardised Radon Index (SRI) is proposed, adapted from Standardised Precipitation Index (SPI) methodologies under development at the University of Northampton (Crockett and Holt, 2010a, b) This has been piloted on a data-set known to contain simultaneous radon anomalies temporally associated with UK earthquakes (Crockett et al, 2006a, b; Crockett and Gillmore, 2010). Different soil properties and conditions at the monitoring sites can result in different nonlinear radon responses to (common) stimuli, such as earthquakes, and this is considered

Radon monitoring and results

Correlation analysis

Standardising radon time-series

Discussion

Conclusions