Spatial relationship between GPS slip and seismic tremor during Cascadia slow slip events

We model GPS deformation and timing of seismic tremor associated with transient deformation in Cascadia to test the hypothesis that tremor and slip occur synchronously but are spatially offset. For the period 2010–2013, we use seismic tremor data with a duration-moment relationship to predict GPS time series and compare them to observations. We find that observed GPS displacements are best predicted when tremor locations on the plate interface are shifted 15 km up-dip of their published epicenter. To test whether the spatial offset of tremor and slip is due to systematic mislocation of published epicenters, we attempt to identify individual sources of tremor using Independent Component Analysis. However, our results are inconclusive. Additionally, our results suggest a moment rate lower than previous studies. We propose that increases in instrumentation have resulted in an increase in recorded tremor giving the appearance of a decrease in moment rate