CONCLUSIONS

It can be concluded that the utility of data recorded at Garm for analyzing location and source parameters of regional earthquakes is good as long as the data are not too much affected by the heterogeneities in the crust. If three different, distinct phases can be identified, the depth can be constrained within 10 km and the distance within 4 km. Identification of two distinct phases and one indistinct phase results still in an accuracy of approximately $\pm 18$ km in depth and $\pm 7$ km in distance. High coda level and indistinct phases, however, destroy accuracy.