Long term air ion monitoring in search of pre-earthquake signals

Abstract

Air ion concentrations at ground level have been measured both indoors and outdoors to address a wide variety of questions, such as estimating air quality or assessing the impact of ions on human health. While most studies rely on discrete measurements, other applications require continuous monitoring over extended periods of time. In this paper, we discuss the challenges posed by long term air ion monitoring within the frame of earthquake forecasting. We first provide a brief state of the art, outlining the main hypotheses advanced to explain anomalies in air ion concentrations that were reported prior to earthquakes, namely the p-hole theory and the radon exhalation theory. We then describe the array of Gerdien capacitors deployed by QuakeFinder in the US and abroad with the goal to identify pre-earthquake variations in air ion data: we discuss the characteristics of these instruments and derive the criticial mobility associated with their parallel-plate geometry. We then present the results of a Parallel Sensor Test (PST) during which positive and negative air ion counters were deployed side by side at increasing separation distances. This test provides insight about the sampling rate at which air ion concentrations should be measured with the QuakeFinder air ion counters.The main processing steps applied to raw air ion data are then described. Particular emphasis is set on the unipolarity coefficient and how it may be used to identify faulty sensors. One of the key issues encountered while monitoring air ion concentrations for extended periods of time is the condensation forming on the electrodes due to increased relative humidity levels. It appears that the relative humidity working range provided by instruments manufacturers is not accurate for long term outdoors measurement and that sensors saturate at relative humidity values lower than the threshold advertised in the instrument specifications. We detail some of the strategies implemented to automatically reject data acquired during episodes of high relative humidity. Finally, we discuss the instrumental improvements that can be made to prevent moisture from forming on the plates of the Gerdien capacitors and that QuakeFinder plans to implement in the near future.