6 Discussion

6.1 3D reflection traveltime modeling

To justify some of the seismic interpretations, we scrutinized the data further. For the most eastern reflection (F1), because it comes near the surface and it is clearly observed in several shot and receiver gathers, we were able to model the reflection traveltime response based on Ayarza et al. (2000), assuming a similar strike as the Wangsukcheon fault, a velocity above the fault of 5000 m/s, and the position where the reflection intersects the surface. A strike of N20E and dip of 60E can explain the reflection traveltime observed in the real data, hence further supporting its origin as being from the Wangsukcheon fault.

6.2 Origin of strong surface-waves back scatterings

Given the strong surface waves back-scattering observed in the data (i.e., Figure 4), it was important to further analyze their properties. The Zerwer et al. (2005) method, as implemented for multifold data by Colombero et al. (2018), was applied to estimate the location of sharp lateral variations in the subsurface and the corresponding maximum affected wavelength. This method shows interesting results for locating lateral variations at different places along the profile. These locations were compared with the bedrock reflection from the landstreamer data and tomography results from the wireless recorders (Figure 5), as well as shot and receiver gathers where surface-waves are clearly back scattered. Two of these surface-wave back scattering sources are particularly strong: (1) where the Chugaryeong fault is mapped and (2) at CMP location 2840. Although no corresponding faults are represented in the geology at CMP 2840, a sharp geological contact between Paleoproterozoic gneiss and Jurassic granite suggests the presence of some tectonic structure that could be the reason for the surface-wave back-scattering energy. The two locations show, respectively, a minimum frequency of 12 Hz and 17 Hz with a maximum affected wavelength of 110 m and 100 m. This distinguishes the scattering to be geological in nature and not due to human constructions such as road and bridge foundations. Similar strong back scattered surface waves are also visible at Pocheon fault surface location, but showing a lower intensity.

6.3 Crossdip analysis and out-of-plane structures

The crooked nature of the profile implies that the trace midpoints are distributed along a 3D zone, allowing evaluation of out-of-the-plane structures and apparent dips for several reflections. This effect was further exploited using a crossdip analysis approach (Bellefleur et al., 1995; Nedimović and West, 2003; Malehmir et al., 2006, 2009; Rodriguez-Tablante et al., 2007; Beckel and Juhlin, 2019). For the reflection (F2) underlying the domed-shaped reflections (Figures 6c and 7), we were able to estimate a crossdip angle of 10 degrees to the north and a true dip of approximately 30 degrees towards NW. The crossdip analysis could only be achieved thanks to the midpoint coverage provided by the crookedness of the profile.

6.4 Fault 3D geometries and seismicity

The information gained from all the analyses were compiled and used to construct 3D surfaces of potential major fault systems along the profile (Figure 8). Based on these surfaces, it is possible to interpret the location and geometry of two of the major fault systems in the area, namely the Wangsukcheon (F1) and Chugaryeong (F3) faults with high reliability. The Wangsukcheon fault dips opposite to what was first expected, especially if considering that it would be a splay fault from the Chugaryeong fault. This implies that the Wangsukcheon fault is likely a separate and unrelated fault system with respect to the two other ones or that it makes a sharp turn as it extends to the northern part of the country (Figure 1). Recent excavation works (Han and Lee, 2019) and historical studies (Kim, 1973) further support our interpretation of the dip direction of the Wangsukcheon fault and the reliability of the 3D reflection traveltime modeling work. The Chugaryeong fault is not imaged as a reflection, as expected for a sub-vertical feature that has also been argued from focal mechanism solutions (Hong et al., 2018, 2021). Nonetheless, there are related features that support the sub-vertical nature of the Chugaryeong fault such as (i) the absence of coherent reflection, (ii) the sharp and important lateral variation visible at the bedrock level in both the landstreamer data and tomography and (iii) the extremely high surface wave back-scattered energy. All these features are observed at the fault surface expression (Choi et al., 2012). The reflections (F2) underlying the domed reflectivity package (Figures 6c and 7) have a true dip of approximately 30 degrees towards the NW. This dip angle was to a certain degree speculated upon by Malehmir et al. (2022) for Pocheon fault but is different from previous suggestions (Hong et al., 2021). Surface projection of these reflections corresponds to the location of strong surface-wave back scattered energy at the mapped position of the Pocheon fault. Interestingly, projecting the reflections towards depth results in them intersecting with the recorded seismicity, suggesting an intersection with the Chugaryeong fault. While this is highly speculative, a possible scenario might be that the Pocheon fault is a splay fault from the Chugaryeong fault system and that the recorded seismicity occurs at their intersection. Malehmir et al. (2022) argued for the same geometry, however they had a much shorter profile, hence their arguments were more speculative. Another possible scenario will be that the F2 reflection is generated by a different fault that becomes steeper close to the surface. Pocheon fault instead will be subvertical, as suggested from seismological focal mechanism inversion studies (Hong et al, 2021), hence not imaged. There may be a third scenario explaining the NW-dipping package of reflectivity underlying the domed-shape reflectivity (F2), namely dykes as they are also interpreted to be present in the domed-shape reflectivity and diffractivity.
Given the curved-shaped nature of the reflectivity overlying these planar sets of reflections, we argue for a thrust (reverse mechanism) fault system associated with these reflections. Thrust faults are known to generate fault-bend folds and this implies the dome-shaped reflectivity might be the result of a growth fault system that was active sometimes (even until now). In this scenario, the reflective package would initially have consisted of sill-type intrusions, which were then folded, and likely also faulted, forming the dome-shaped reflectivity observed in the central part of the study area (Figure 7). Assuming this interpretation scenario is correct, a direct implication is that both the Chugaryeong and F2 faults might be active. Given the opposite dip direction of the Wangsukcheon, it is likely that any seismicity along this fault should be separately looked at, although the fault appears to be inactive in terms of seismicity recorded in the area.