JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 86, NO. B3, PAGES 1725-1730, MARCH 10, 1981 Relationship of the 1979Southern California Radon Anomaly to a Possible Regional Strain Event M. H. SHAPIRO, • J. D. MELVIN, T. A. TOMBRELLO,AND M. H. MENDENHALL W. K. Kellogg Radiation Laboratory,California Institute of Technology, Pasadena, California91125 P. B. LARSON Geological andPlanetary Sciences, California Institute of Technology, Pasadena, California 91125 J. H. WHITCOMB CIRES, University of Colorado, Boulder, Colorado 80309 , During the second haft of 1979,anomalously high emanation of radon was recorded at two stations of the automated radon-thoron monitoring network operated by the W. K. Kellogg Radiation Laboratory of the California Instituteof Technology. The two stations exhibiting major anomalies, Kresge and Dal- ton Canyon, are located approximately 30 km aparton the frontalfault system of the Transverse Ranges of southern California. At Kresge the anomalybegan on June 21, 1979,and continued throughDecem- ber 1979. At Dalton Canyonthe anomalystarted about 3 weeks later and alsocontinued through Decem- ber 1979.At both sites the anomalous levelsof radon decreased (but did not return entirely to normal values)shortlybeforeOctober 15, 1979.During the week of October 15, 1979,a 6.6-M earthquake oc- curred about290 km to the southeast of the two stations, and later in that week,earthquakes of magni- tude 4.2 and 4,1 occurred at Malibu and Lytle Creek. The latter two events were within 60 km of the monitors. A radon-thoron monitor at Lytle Creek recorded no long-termanomaly but did recorda sharp spikelike decrease in the radon level on October 13, 1979.Coincident with our observations of anoma- lous radon levels,other investigators have reportedanomalies or suspected anomalies in severalother geodetic, geophysical, and geochemical signals from the same general region. The rapid temporal devel- opmentof several of the anomalies together with the large area over which they were observed suggests that a large-scale strain event took place which may have been responsible both for the widespread anomalies and for the seismicity that occurred in the regionsubsequent to the onset of the anomalies. INTRODUCTION Since the 1965 Tashkent earthquake there have been a number of reports of radon anomalies preceding moderateor large earthquakesfrom the USSR [Sadovsky et at., 1972], China [Li et at., 1975],and Japan [ Wakita et at., 1980].Almost all of these reports have been based on the observationof changes in radon levelsin groundwater. More recently, in the United States, severalgroups have been monitoring radon in soil gas[Mogro-Campero et at., 1980; King, 1980; Birchardand Libby, 1980] andin groundwater [Shapiro et at., 1980a; Tal- wani et al., 1980; Teng, 1980] in an attempt to observe pre- c .ursory changes in radon levels. While several of the reported precursory radonanomalies havebeenrecorded within a few tens of kilometersof the subsequent epicenter,there have been a number of reports of precursory radon anomaliesat large distances from subsequent epicenters.For example, Teng [1980]tabulated data from the Kutzan stationin China, where in a 4-year period, radon spikes of shortduration were observed several days before eight earthquakes ranging in magnitude from 5.2 to 7.9. All of these earthquakes occurred on a Y-shaped fracture zone. The nearestevent with a pre- cursoryanomaly was 54 km from the station,while the fur- thestoccurred 345 km from the station. During the samepe- riod one radon spike was observed with no subsequent earthquake, and for one 6.5-M earthquake 420 km distanton • Also affdiated with thePhysics Department, California State Uni- versity,Fullerton, California 92634. Copyright ¸ 1981 by the American Geophysical Union. the same fracture zone no precursory anomaly was observed. Such reportsof supposedly precursory radon anomaliesat great distance from subsequent earthquakes often have been received with considerable skepticism, particularly in the United States, since the haft-life of a radon atom is too short (3.8 days) to permit its movement more than several tens of metersfrom its point of productioneven if active subsurface transport mechanisms are invoked[Tanner,1978]. In addition, during the shorthistoryof the radon monitoring effort in the United States there have been only a few previous reportsof precursory radon anomalies, and generally,thesehave been associated with nearby earthquakes.In this paper we report the observation of a radon anomaly which coincided with sev- eral other geophysical and geochemical anomalies and which appears to have beenassociated with an earthquake some 290 km from the site of the radon anomalies. These observations are discussed with reference to two recently proposed models which would account for much of the anomalous data ob- servedin southernCalifornia during 1979. EXPERIMENTAL PROCEDURE A full description of the automatedradon-thoronmonitors used in the California Institute of Technology radon mon- itoring program hasbeengivenby Melvin et al. [1978].Radon data from the Kresgesite,which is locatedin the San Rafael Hills of Pasadena (34.089øN, 118.103øW), have beenobtained from two monitors.The first prototype unit monitored radon and thoron from a 24.1-m-deep static borehole in granitefrom April 1977to March 1979.Data from this periodof operation Paper number 80B1527. 0148-0227/81/080B- 1527501.00 1725