Figure 5. Optimal initial velocity model inverted from the VELEST package. (a) Starting input models (gray lines and the specific velocity values are shown in Table  2 ) with different gradients and the inverted models (black lines) from VELEST package. (b) The average velocity (blue line

Figure 5. One‐dimensional velocity models. Blue is used for HYPOINVERSE locations and regional moment tensor inversion, based on Pechmann et al. , 1995 . Red is our preferred model from VELEST inversions used for single‐event VELEST and relative hypoDD relocations. Based on VELEST inversions

Figure A2. Initial locations for 715 out of 725 local earthquakes (sized by magnitude) near the DCPP (star), used as input into a second VELEST location run with station corrections; Figure  6 shows the output locations after the second VELEST run. For the 265 catalog earthquakes detected

Figure 5. (a) The suite of eight trial velocity models used as input to VELEST to explore the model space for the local velocity model. Models labeled 1–3 consist of nine layers with heterogeneous P ‐wave velocities, models 4–6 consist of nine layers of homogeneous P ‐wave velocities, and model

Figure 6. (a) The suite of eight trial velocity models used as input to VELEST for exploring the model space of the regional model of western Montana. Models 1–3 consist of fifteen‐layer heterogeneous P ‐wave velocities, models 4–6 consist of fifteen‐layer P ‐wave velocities, and model 7

Figure 4 Contour maps of VELEST station delays: (a) P-delay and (b) S-delay. Inverted triangles denote station locations.

consistent with aftershocks. (c) VELEST absolute locations with a VELEST‐derived 1D velocity model (Fig.  5 ). Otherwise as panel (b). (d)  hypoDD relative locations of 1401 events with the VELEST 1D model. Colors indicate temporal distribution, as in the upper left. Note: Tight relative locations from

Figure 2. P-wavespeed versus depth for VELEST models ( Table 3 ; preferred model in red; thin lines represent the suite of inversion solutions). Northern California Seismic Network (NCSN) model ( Oppenheimer et al., 1993 ) is shown in green and the Miller and Mooney (1994) model is shown

Figure 8. (a)  P ‐wave travel‐time corrections obtained from VELEST. The reference station (NPW) is depicted by a gray diamond. (b)  M L station corrections. The stations discussed in the Discussion section are labeled.

Figure 8. Aftershock locations from VELEST for 1274 events from the final 1D velocity inversion with station corrections. Seismic stations, triangles. Clusters of hypocenters within the rectangular boxes are included in the corresponding cross sections and are referred to in the text as clusters

Figure 4. Velocity models for P waves obtained with VELEST for Moho depths ranging from 33 to 42 km. Two different starting velocities in the upper and lower crust and mantle were used and marked as MIN and MAX (see Table  3 ). Obtained rms values are classified by colors, ranging from light

Figure 5. P ‐ and S ‐wave velocity model obtained with VELEST for the study area.

Figure 3. Map showing the 268 earthquakes relocated by the HYPOINVERSE/VELEST steps (white dots), and the 427 events that were relocated up through the hypoDD double‐difference algorithm (black dots). About 85% of the events locate on the Pa–NA plate boundary within the GoC . Note

Figure 3. Distribution of the ray paths used in VELEST ( Kissling et al. , 1994 ) shown with the polygon for the selected catalog earthquakes. Dashed lines show the outlines of the Rhenish Massif.

Figure 3. A summary of the velocity models used in this study. The model of McLaren and Savage (2001) is used as the starting model for a VELEST ( Kissling et al. , 1994 ) inversion for the best 1D model. A gradient approximation of the VELEST model is used as the starting model for the 3D

Figure 3. (a) Solution of seismic velocity in the study area from simultaneous inversion in the Velest program. The red line is the best solution that converges from various models, as represented by the gray line. (b) The root mean square (rms) graph of Velest processing from best model denotes

Figure 3. (a) The probability of the layer velocity obtained for the top 100 models obtained from VELEST with low root mean square (rms) error. The Moho is observed at a depth of 30 km. The proposed minimum 1D model is shown as solid line whereas the velocity models after Rao et al. (2011

Figure 2 RMS residual vs. number of earthquakes before VELST (blue bar) and after VELEST (red bar).

Figure 2. Procedure to detect stations with systematic errors in travel-time data with VELEST.

Figure 3. Flowchart of the process used to determine the 1D velocity model at each site using VELEST.