Constraints on crustal attenuation and three-dimensional spatial distribution of stress drop in Switzerland

Abstract

We employ two different methods to estimate attenuation-corrected source spectra and calculate Brune stress drops of small to medium earthquakes in Switzerland and the surrounding region. We test trade-offs between source and travel path contributions by comparing results between a data-driven and a model-dependent method of distinguishing between the two. Systematic differences between results from both methods can best be reconciled by assuming intrinsic whole-path attenuation to be frequency dependent. A two-parameter grid search based on a power-law Q( f ) function reveals trade-offs that prevent us from quantifying a more exact frequency dependence. However, a comparison of the two-source parameter inversion methods with synthetic tests provide evidence for a non-negligible frequency dependence of Q between 1 and 30 Hz. Consideration of these implications on the attenuation model, as well as consideration of lateral and vertical variations of velocity and quality factor reduces the scatter of the obtained stress drop estimates. Synthetic tests confirm that both methods are able to robustly resolve lateral variations of Brune stress drop with quantifiable uncertainty estimates. Resulting lateral variations show reduced stress drop along the Alpine deformation front. This pattern points to tectonic causes and may be due to variations in differential stress.