11 Nature and Spatial Probability of Future Rock Deformation in Each Domain using Example Locations

Transcription

1 11 Nature and Spatial Probability of Future Rock Deformation in Each Domain using Example Locations Two approaches to considering the probability of future rock deformation in Kyushu have been developed. Neither of these approaches is mutually exclusive and, ideally, both should be integrated into a single methodology. The first approach is based on the precept that the past is the key to the future and that we may therefore use current rates of activity to assess the probability of exceedance of various strain rates as a proxy for future rock deformation. Simple calculations relating earthquake energy to scaling of fault slip to earthquake magnitude, and converting that to strain rate, suggests that an annual strain rate of 4 nanostrains/a within a unit area of 5 x 5 km is consistent with rock deformation that may occur during the equivalent of 1 m of surface displacement (equivalent to a single magnitude 7 earthquake) in a period of 100,000 years. This calculation of course assumes that all strain calculated for a particular 5 x 5 km area from the various rock deformation datasets will be converted to a discrete faulting event across a discrete fault every 100,000 years, which is probably far too simplistic an approach. However, given various assumptions, threshold hazardous strain rates can be defined for specific rock deformation scenarios of concern, once those scenarios are decided upon. Plots showing the probability of exceedance of various strain rates for example locations in Kyushu (Figures 11.1 and 11.2) enable comparisons between locations, and give us some idea of how likely a given example location is to exceed a certain level of strain. Whether the various strain rates constitute a risk to a repository depends on the repository design and the manifestation of the rock deformation at those locations. The deformation may range from concentrated displacement on a single fault trace to widely distributed, diffuse, fracturing within a 5 x 5km area enclosing the location. The second approach is to take note of the rapid geological evolution of the Kyushu region and consider the possibility that future rock deformation may evolve in both space and time, and that while past and current rates of deformation capture this evolution to some extent, we can also qualitatively assess what sort of evolving activity may affect the example locations. Key features or processes we observe to be driving the rapid evolution, and their impacts in Kyushu are: (i) (ii) (iii) (iv) Roll-back at the Nankai Trough, which may induce migration of arc volcanism southeastward Southwestward migration of the Kyushu/Palau Ridge (and its subducted part) beneath Kyushu because of the oblique angle of subduction and the orientation of the volcanic chain comprising the Kyushu/Palau Ridge. We suggest that an active left-lateral shear zone in southern Kyushu (revealed by GPS and seismicity) occurs due to the impingement of the Kyushu-Palau Ridge on the subduction margin (see discussion in Section 9). It has also been suggested by previous studies that uplift in the forearc above the ridge subduction point is related to Kyushu-Palau Ridge subduction. It is also possible and localisation of heat flow and Aso-type volcanism in the arc is related to the position of the Ridge (see Section 7). Thus, we could expect left-lateral shear of southern Kyushu, uplift in the forearc, and localisation of certain types of volcanism to migrate also. The northeast limit on extension associated with the Okinawa Trough is adjacent to the western end of the Beppu Shimabara Graben (BSG) in the vicinity of Nagasaki and Kumamoto. It is unclear whether this zone of extension continues to develop, but it could be increasing the likelihood of normal faulting and volcanism in that area. Extension in the Kagoshima Graben and volcanism such as at Sakurajima are geologically recent phenomena in southern Kyushu. Future evolution and development of this zone may be expected. Future steepening of subduction at the Ryukyu Trench as a consequence of slab roll-back will encourage migration of the volcanism in the Kagoshima Graben toward the east and southeast. 126

2 Figure 11.1 (continued on next page): Cumulative strain rate plots for the 11 example locations. EQ WGTD AV is equal weighted average. See Figure 6.4 for location of the 11 example locations. 127

3 Figure 11.1 (continued): Cumulative strain rate plots for the 11 example locations. EQ WGTD AV is equal weighted average. See Figure 6.4 for location of the 11 example locations. (v) (vi) At present the volcanism and faulting associated with the Kagoshima Graben does not extend further northeastward than Kirishima volcano, with a marked gap in volcanism in central Kyushu between Kirishima and Aso volcanic centres. In normal situations we would expect that the volcanic arc would continue to the north approximately located above the 100 km depth contour on the subducted Philippine Sea Plate. We are not aware of any current data to explain this gap in volcanism and thus should consider the possibility of new volcanoes developing in that area. In the Back-arc domain of northern Kyushu and offshore Japan Sea, large earthquakes are occurring on reactivated northwest- and northeast-striking faults. There may also be correlation of these structures to some of the monogenetic volcanism. For example, stress perturbations that arise at fault intersections may be preferred locations for future small volume monogenetic events. The backarc province hosts a wide variety of volcanism that is dominated by processes in the mantle and the mantle wedge rather than subduction. Little is known of the processes leading to irregular volcanic flare-ups in the Back-arc domain and there is large uncertainty is forecasting future activity. 128

4 Based on the features and processes noted above we discuss current strain rates estimated for these example locations, and offer some suggestions of possible future rock deformation evolution at or adjacent to each of the example locations in the various domains defined in the Kyushu Case Study. To aid with the comparison between locations, Figure 11.2 shows all the probability of exceedance curves on one plot. Figure 11.2: Combined equal weighted average cumulative strain rate plot for the 11 example locations, colour-coded by tectonic domain. See Figure 6.4 for location of the 11 example locations Back-arc domain Example Location 1. Located on an active fault with a low slip rate of 0.1 mm/a. There is a general good agreement on strain rates between all 3 datasets: current GPS strain rates are low-moderate, as are seismicity and active fault strain rates (Figure 11.1). Future deformation can be expected in the form of sinistral strike-slip surface faulting associated with earthquakes up to about M 7.5 (see Section 10). Because the fault is reactivating an earlier structure the fault zone complexity (the process zone) can be expected to be wider than on a long-lived, well-developed, fault. Example Location 2. Located about 20 km from a low slip rate (0.1 mm/a) active fault. Current GPS strain rates are moderate to low, and seismicity strain rates are very low (Figure 11.1). Future deformation is a possibility but it is unlikely the process zone of future faulting on known active faults in the region (20 km away) will extend to this location. Secondary faulting or fracturing of suitably oriented bedrock faults and fractures is possible in the event of a large nearby earthquake. Example Location 11. Located in southwestern Honshu in the southern part of the monogenetic Abu volcanic field. In the region there are both northeast- and northwest-striking faults, although the closest known active fault is more than 20 km to the west. The west Tottori earthquake of 2000 occurred approximately 200 km to the northeast in a similar tectonic environment. Current GPS strain rates are moderate, and seismicity strain rates are 129

5 very low (Figure 11.1). Reactivation of bedrock structures in this region as strike-slip faults is possible considering other active faults in the region. There is a possible structural control on the location of volcanic edifices of the Abu volcanic field and, at this location, possible future co-location of volcanism and faulting should be assessed Extensional domain Example Location 3. Located west of Aso Volcano, within the BSG, and about 12km north of a prominent normal fault extending westward from Aso Volcano. Thick, young pyroclastic deposits cover the area, which may have obscured some faulting near to the location. Current GPS strain rates are high, and seismicity strain rates are very low (Figure 11.1). Further development of the normal faults can be expected in the region. Proximity to Aso Volcano suggests that ground deformation associated with future volcanism is also probable. Example Location 10. Located about 30 km SW of Aso Volcano near the southeast margin of the BSG, and within 5 km of a moderate slip rate normal fault (~0.9 mm/a). Thick, young pyroclastic deposits cover the area, which may have obscured some faulting near to the location. Current GPS and active fault strain rates are high, and seismicity strain rates are very low (Figure 1). Further development of the normal faults can be expected in the region. Proximity to Aso Volcano and associated connected edifices suggests that ground deformation associated with future volcanism is also probable. This location is possibly in the region where future Aso-type volcanism and related tectonism may concentrate with the migration of the Kyushu Palau Ridge subduction point intersection with the BSG. If this process is significant in governing the location and migration of a nexus of volcanism, then Aso-type volcanism could migrate to a location 10 position, million years from now. Current strain rates at Location 10 are among the highest observed in the study. Example Location 4. Located in northeast Kyushu at the eastern margin of the BSG. The location is on the flanks of a volcano but north of the current extent of normal faulting associated with the BSG. Geodetic strain rates are high and very similar to other locations in the BSG (Figure 11.1). Seismicity strain rates are very low (Figure 11.1). Future hazard may diminish at this location as the shallowly subducting Shikoku Basin crust encroaches on this region and shuts down any remaining extensional deformation, over the next 1-2 Ma Forearc domain Example Location 5. Located in the Forearc domain immediately east of the boundary with the Extensional domain and about 15 km east of Aso Volcano. The location is between active traces of low slip rate faults (~0.1 mm/a), and bedrock geology in the area has many steep faults oriented parallel to the domain boundary between the Extensional and Forearc domains. Current geodetic strain rates are high and seismicity strain rates are low (Figure 11.1). Future deformation at this location might occur as a result of reactivation of bedrock structures that evolve the current unlinked fault strands into a more continuous feature. If trench roll-back at the Nankai Trough is significant process then it is likely the southeastern margin of the Extensional domain will also migrate into the current Forearc domain. Deformation at the location could therefore be expected to switch to normal faulting and increase in rate. The location is likely to be heavily inundated by ash fall in the event of a major eruption from Aso Volcano. Example Location 6. Located near the southeast coast of Kyushu approximately 15 km north of the nearest known fault at the margin of the Miyazaki Plain. Current GPS strain rates are high and seismicity strain rates are low (Figure 11.1). However, it is possible that the GPS strain rates are artificially high here due to choice of subduction interface coupling models (see discussion in Section 9). Coastal uplift rates of about 0.2 mm/a have been noted nearby and rates up to 0.7 mm/a further south along the Miyazaki coast. If these rates extend inland to this location this could result in hundreds of metres of uplift over long time periods, and represent a significant rock deformation. 130

6 Example Location 7. Located in an area about 30 km distant from the nearest active fault and about 30 km from the BSG. The location is above the 100 km contour on the subducting oceanic plate but in an area of no current or past volcanism. There is strong northeast striking bedrock structure in the area. Current GPS strains are high and seismicity strain rates are low (Figure 11.1). This location is near to the junction of high left-lateral shear strain cross-cutting Kyushu (as seen from the GPS velocities: see Section 9) and the southeast boundary of the BSG. Given that this location is adjacent to the intersection of the left-lateral shear zone with the BSG, we cannot rule out migration of future rock deformation into this region. Obtaining improved mantle structure data may indicate reasons for the gap in arc volcanism. This would be a key future requirement if a location such as this were to be considered as a potential repository site Southern Arc domain Example Location 8. Located about 15 km to the east of faults marking the eastern boundary of the Kagoshima Graben. The location is on the southern end of one of the northeast striking, low slip rate (~0.1 mm/a), faults of the region, and is just to the south of the left-lateral shear zone cutting across southern Kyushu (from GPS). Current GPS strain rates are high, active fault strain rates are moderate, and seismicity strain rates are low (Figure 11.1). Future rock deformation may be characterised by rare surface rupture of the local active fault, or speculatively by significantly increased rates of deformation if the Kagoshima Graben is migrating with the roll-back of the subducted plate, and by southward migration of the left-lateral shear zone cross-cutting southern Kyushu which may be driven by subduction of the Kyushu-Palau Ridge. Example Location 9. Located about 15 km west of the western margin of the Kagoshima Graben. The location is within the former volcanic arc (before it migrated eastward to the Kagoshima Graben area) (Yamaji et al., 2003). The location is about 10 km south of northeast striking normal cutting across the old arc volcanoes and is just to the south of the left-lateral shear zone cutting across southern Kyushu (from GPS). Current GPS strain rates are high, but slightly less than location 8, and seismicity strain rates are very low, also less than location 8 (Figure 11.1). Future rock deformation at this location might be less likely than at location 8 because it appears to be in the trailing position of the locus of volcanism in the contemporary Kagoshima Graben. The characteristics and possibility of migration of the faulting located about 10 km to north of the location would be important to understand in further investigations at this locality. In particular, southward migration of the left-lateral shear zone cross-cutting southern Kyushu which may be driven by subduction of the Kyushu-Palau Ridge could be important in future rock deformation at this location. 131