| Summary: | From June 1978 until late 1980 radon emission from the Leirhnjukur fumaroles was monitored within the Krafla caldera of north Iceland where episodic volcanism is occurring. Frequent sampling of the fumaroles shows that no easily identifiable short-term radon precursors occur in the days prior to subsidence of the caldera, despite an observed increase in microseismicity preceding deflation. Following the onset of subsidence, however, the radon emission of the fumaroles gradually increases and reaches a maximum 3–6 days later. The radon in the fumaroles is assumed to be transported from depth by steam and noncondensing gases that slowly escape from the geothermal water table. The cause of the co-episodic increase in radon emission appears to be a temporal rise of the water table driven by fissure closure resulting in an abbreviated transport time for radon to the surface fumaroles. Furthermore, the closing of the fissures appears to cause a transient increase in the velocity of transport, making the shape of the anomalies broader and higher than is predicted from a change in the water level alone. Changes in radon emission also coincide with fluctuations in fumarolic activity and permanent changes in the level of geothermal water that occur during periods of uplift. © 1981 by the American Geophysical Union. Received April 4, 1981; revised July 24, 1981; accepted October 2, 1981. The author is indebted to the following colleagues, who kindly made unpublished data available for this study: P. Einarsson at Science Institute, University of Iceland, provided the earthquake data. A. Bjornsson and V. Stefansson at the National Energy Authority of Iceland provided the land elevation and the water level data. H. Tryggvason, also at the National Energy Authority, provided the data on the fumaroles and also collected the radon samples, often under difficult weather conditions. Valuable assistance from participants in the LDGO-Iceland radon project, J. G. Goddard, G. Aradottir, and S. E. Palsson is appreciated. The ...
|
|---|