Figure 1:  Model of a basic fault and different source locations.
Left: source outside the fault, middle: source at the FZ boundary, right: source centered in the fault.
The source depth is 6 km. the width of the FZ is 270 m. The size of the model is 5 km in x direction and 7.5 km in z direction.

Figure 2:  Seismograms for the models shown in Figure 1.
The y-component of velocity seismograms are drawn to scale.
Left: source outside the fault, middle: source at the FZ boundary, right: source centered in the fault.
The grey rectangle denotes the stations above the fault.
The maximum amplitude of each trace for both the direct (dotted line, marked with S) and the trapped wave
(solid line, marked with FZ) are displayed on top.

Split Fault Model

Figure 3: Left: Split-fault model with the source shifted laterally.
Middle: Seismograms of receiver line L2 situated on top of the forepart of the fault. For the receiver positions see here.
Right: Seismograms for receiver line L4 situated on top of the rear part of the fault. The seismograms of the receiver line with the strongest FZ waves on each fault segment and the maximum amplitude distributions of all six receiver lines are shown.
Even the stations on profile L1-3 of the fault show clear FZ wave arrivals.

Fault With Small Scale Scatterers

Figure 4: Left: model with small scale inhomogeneities.
The FZ properties vary from -30% to 0% with respect to the surrounding quarterspaces.
Middle: seismograms for a reference model with uniform seismic FZ properties which corresponds to the average of the inhomogeneous model. Right: seismograms for a fault with small scale inhomogeneities. 

Gradient vs. Non-Gradient Fault Model

Figure 5: Left: models with a fault widening towards the surface and a low velocity layer on top.
Model A: constant seismic properties at depth. Model B: velocities and density increase with depth.
Middle: seismograms for Model A. Right: seismograms for model B.