Vinsamlegast notið þetta auðkenni þegar þið vitnið til verksins eða tengið í það: http://hdl.handle.net/1946/34507
The monitoring of geothermal areas in Iceland is very important, as these are one of the main sources of electricity and hot water. For this purpose, Iceland Geosurvey (ISOR) runs seismic networks for the power companies in most of Icelandic geothermal areas, including Krafla.
The short time average / long time average (STA/LTA) algorithm used for earthquake detection can be adjusted by certain user-set parameters. With the aim of improving the monitoring of the Krafla geothermal area, parameter tuning in the SeisComP3 detection system was made by brute force calculation of thousands of possibilities. This was done by manually checking the traces and writing the picking times for each of the stations for selected time periods, and comparing it to the brute-force calculations carried out using a python script. A set of optimal parameters for the Krafla geothermal area was then selected from the top results of each station, also considering that similar parameters between the stations are advantageous for simplicity reasons.
The results showed that even though the selected set of parameters has been enhanced the detection rate is considerably lower than with ISOR defaults. This can be explained by a limitation of minimum P phases that SeisComP3 has, and as a consequence events that are only detected in less than 6 stations are omitted. The conclusion is that the selected set of parameters proves to be unpractical to implement and under the current circumstances ISOR default parameters are more appropiate. However, significant improvements could be done if SeisComP3 limitation could somehow be bypassed.