Task 2: Revive the radon sampling program in South Iceland

Start: March 1996 (month 1)
End: April 1998
Responsible partner: UICE.DG

This part of the project is funded by the SEISMIS project in Iceland. The start of the reviving of the radon sampling program has been delayed due to delays in building the necessary instruments (Task 1). Regular sampling of radon and analysis using the new instrument will start during the summer, funded by the SEISMIS project in Iceland.

In spite of the delays in finalizing the construction of the LSC instrument for the radon analysis and resulting delays in starting the regular radon monitoring, significant work has been carried out, cost by the contractor, which is a good preparation for the new regular radon sampling program. This work was concentrated in two areas:

• Analysis of radon time series collected in South Iceland 1977-1993. These represent some of the longest time series available. The results have been reported in several publications [45,,,] and a manuscript will be submitted to a refereed journal within the next month [].
• Development and testing of some of the physical concepts behind the improved LSC apparatus. This work was funded by other projects. This work has been documented by Theodórsson [84] and in a recent book on weak radioactivity [85].

The work on analyzing the 1977-1993 radon time series can be shortly summarized as follows:

During the period 1977-1993 radon was monitored in water samples from several geothermal drill holes in the South Iceland seismic zone. The 3-16 years long time series are generally smooth functions with sharp anomalies or excursions superimposed on them. The anomalies show a distinct relationship with earthquake activity in the region. About 35% of all measured anomalies are related in time to earthquakes. More than 30% of all seismic events satisfying certain magnitude-distance criteria (M $\geq$2.0 and M $\geq 2.4 \log{D} - 0.43$) are accompanied by anomalous excursions in the radon time series. Seven events were preceded by anomalies detected at more than one station. One event was accompanied by anomalies at 5 stations. Majority of the anomalies are positive and occur before the respective seismic events. There is a 38% probability that a measured, positive anomaly is immediately followed by an earthquake. The efficiency of the sampling stations at detecting earthquake-related anomalies is found to be positively correlated with the average radon activity of the water, which is related with the radioactivity of the host rock. The success rate of a monitoring network of radon for the purpose of earthquake forecasting can most likely be improved from these figures by increasing the sampling rate, refining the definitions of what constitutes a radon anomaly at each sampling station, and a careful selection of sampling sites. Several anomalies were found that are related to the eruptions of the nearby volcano Hekla in 1980 and 1981.