Time series regression techniques and applications to radon data

Abstract

There have been many studies in the last few years suggesting that radon emissions from groundwater or rock may give useful information for the prediction of earthquakes. Klusman et al. (1981), Shapiro et al. (1981, 1982, 1984, 1985), King (1980, 1985) and Teng (1980) suggested that changes in radon emissions might be expected prior to large earthquakes. However other meteorological factors such as atmospheric pressure, air temperature, groundwater level, rainfall, earthtides etc may affect the variation of radon emissions. In this thesis I study radon emissions from groundwater in a deep well in Kelburn from December 1982 to February 1985. An attempt is made to investigate the relationship between radon emissions and other meteorological factors such as atmospheric pressure, air temperature, rainfall, earthtides and groundwater level in the Wellington region. Time series techniques in both time and frequency domains are used. The study was first suggested by Dr. Russell Robinson of Geophysics Division, DSIR and has been under the supervision of Prof. David Vere-Jones and Dr. Peter Thomson of the Institute of Statistics and Operations Research, Victoria University of Wellington.